I’ve chosen to close Part One of this book with a frontal assault on the juiciest bovine of all: the unassailable belief that salespeople should be paid commissions. And while I’m at it, I’ll take aim at bonuses, targets and other artificial management stimulants.

A litmus test

This discussion is important for two reasons.

First, commissions and their bedfellows will definitely handicap the performance of the reengineered sales environment I’ve gone to great lengths to describe.

Second, this discussion will force us to confront the significant implications of Sales Process Engineering: both locally and organization wide.

If you are brave enough to follow in the footsteps of our quiet revolutionaries it’s critical that you truly appreciate the essence of SPE. It’s not enough to believe that SPE will work; you must also understand – at the most fundamental level – exactly why it will work. (And if you don’t, it almost certainly won’t!)

So, I’m proposing that you use the (emotionally-charged) question of salespeople’s commissions as a kind of litmus test. If, by the end of this chapter, you are comfortable that there is no place for commissions in a reengineered sales environment, it’s safe for you to proceed.

If, however, this conclusion still does not sit comfortably with you, it makes more sense to treat this book as an exercise in creative thinking (and leave your sales function well alone).

When commissions make sense

At its most fundamental level, SPE involves the transitioning of the responsibility for sales from autonomous agents to a centrally-coordinated team.

When sales actually is performed by autonomous agents, it does makes sense to pay these agents on a commission basis (a percentage of the revenue they generate).

So, if we imagine a computer hardware manufacturer that sells desktop and notebook computers to consumers, via big-box retailers; it’s clear that these arms-length retailers should be paid on a commission basis.

And, if we think about this example, we can identify two conditions that accord well with commission-based pay:

1. These retail agents sell from stock – meaning that there is no requirement for them to interact with the hardware manufacturer on a transaction-by-transaction basis (which certainly would not be the case in an engineer-to-order environment)
2. These retail agents are truly autonomous – they march to their own drumbeats (and they own the relationship with the ultimate customer)

But what happens to the case for commission-based pay when these conditions are not in place?

As we discussed in Chapter 2, when we transition from a make-to-stock to an engineer-to-order environment, the case for autonomy becomes weaker. Increasingly, the performance of the organization as a whole becomes more a function of the quality of integration between sales and production.

And, because autonomy and teamwork are polar opposites, as the case for autonomy becomes weaker, we reach a point where we have to make a clean switch from one to the other (there’s simply no such thing as an autonomous team member!).

The wrong question

We now arrive at the critical question. We should not begin this discussion by asking: does commission make sense? Rather, we should ask: should we sell via autonomous agents or via a centrally-coordinated team?

Once we answer this question, our position on commissions becomes obvious.

Commissions: the case against

In order to understand why, let’s briefly revisit the history of manufacturing.

There was a time (before the industrial revolution) when almost all labor was paid on a piece-rate. Piece-rate pay is the manufacturing equivalent of commission. Rather than being paid in units of time, a piece-rate worker is paid in units of output. (A sewer, for example, might receive 20c for each garment processed.)

Today, however, piece-rate pay is almost extinct. (And, I suspect by now, you have a good idea why!)

What happened is that management discovered that, as the complexity of the environment increased, there was a critical threshold beyond which scheduling decisions had to be made centrally.

Of course, beyond this threshold, piece-rate pay had to be eliminated because it drove workers to work as fast as possible and not to subordinate to the schedule. (Remember, because of the combination of dependency and variability you never maximize the output of a system by maximizing the rate-of-work of each system resource.)

Commissions (or any kind of performance pay) are inappropriate in the reengineered sales environments described in this book for exactly the same reason that piece-rate pay is now inappropriate in manufacturing environments.

And this conclusion does not just apply to the sales function in isolation. As we discussed in Chapter 4, in many organizations it is not healthy for sales to be the organizational constraint. So, in these cases, irrespective of the structure of the sales function, the organization as a whole will perform better when sales is not operating at 100% utilization.

I wish this could be my last word on that subject. However, there’s a number of persistent objections to my position that we must first put to bed:

1. Ours is a mixed environment: salespeople are not fully autonomous – meaning that a mix of salary and commission is justified
2. Even if we don’t need the compensation plan to determine salespeople’s rate of work, we still need performance pay to maximize salespeople’s quality of work (in other words, without commissions, what would motivate salespeople to actually sell?)
3. Commissions enable us to mitigate against the uncertain nature of salespeople’s performance and keep costs under control
4. The theory may make sense, but good salespeople will simply not be prepared to work in an environment without commissions

The fallacy of a mixed environment

I’ve heard many executives argue that it’s beneficial for their salespeople to be partly autonomous. But I’ve never heard anyone argue that it’s beneficial for salespeople to be partly team members.

Perhaps it’s because the latter phrasing exposes the folly of this position!

I’ve already stressed that it is impossible for salespeople to be team members and autonomous agents at the same time. However, an astute reader might argue that this is possible in theory (if not in practice).

Your salespeople can be capable team members and operate autonomously if (and only if) the rest of the organization has the capacity to subordinate to individual salespeople.

At first glance, this condition may not appear to be particularly onerous. However, when we consider the enormous variability in salespeople’s output, we recognize that effective subordination would require a huge amount of redundancy in customer service, engineering and production. (Remember, we’re considering true sales here, not repeat transactions.)

The fact that this is commercially unrealistic in most organization tends not to stop management from pursuing a mixed sales environment. And, the consequences are as unpleasant as they are predictable:

1. Management encourages salespeople to operate autonomously
2. Salespeople proceed from the assumption that more sales is always better (the tacit assumption is that the rest of the organization can keep up)
3. On average, the sales team as a whole sells less than the organization has the capacity to produce
4. However, because new accounts are won infrequently (after all, salespeople spend the greater majority of their time processing transactions), the load on the rest of the organization is irregular
5. On the (not infrequent) occasions that customer service, engineering or production does not have the capacity to honor commitments made by salespeople, on-time performance tends to be compromised (although, the commitments that have been made to new accounts are often met, at the expense of existing ones)
6. Periodically, management attempts to improve the financial performance of the organization with additional incentives and special promotions
7. These incentives tend to increase the lumpiness of the deal flow – meaning that, over time, peak sales increase, at the expense of average sales

The bottom-line is that contradictions cannot persist indefinitely. Your salespeople cannot be both autonomous agents and team members. They cannot be responsible only for sales outcomes and simultaneously be expected to attend sales meetings and maintain the organization’s CRM. And customers cannot belong to both salespeople and to your organization.

Commissions and the quality of work

If salespeople don’t have the opportunity to earn commission, then why would they sell?

I wish I had a dollar for every time I’ve been asked this question by an incredulous executive. You would think the onus should be on the defender of performance pay to present an argument.

After all, receptionists answer the phone when it rings, in spite of the fact that they receive no incremental pay. Your financial controller does a good job of paying bills on time, in spite of the fact that they receive no rebate on each check signed. And even senior executives perform important tasks, absent special incentives (I’m assuming that no one is paying you to read this book!).

Why should salespeople differ from almost every other worker on the planet?

The answer to the question four paragraphs above is simple: absent the opportunity to earn a commission, salespeople will still sell because they are salespeople. (Just as receptionists answer the phone because they are receptionists.)

I often wonder if those executives who ask that question are really enquiring into the motivation of their team members or if they are providing an (unsolicited) insight into their own pathology!

In Drive, his excellent best-seller, Daniel Pink presents a powerful case against performance pay. His conclusion – backed-up by many experiments from the social sciences – is that external rewards retard the performance of knowledge workers and have a positive effect only in situations where workers are performing mindless, repetitive tasks.

In other words, if your team members are responsible for activities any more complex than licking stamps, the work itself is their reward.

Interestingly, Pink’s conclusion points to an interesting defense of performance pay in the traditional sales environment. Consider these two points:

1. In most environments the volume of sales appointments has a far greater influence on sales output than the qualitative performance of the salesperson
2. In almost all environments, salespeople generate their own appointments as a result of mindless and repetitive prospecting activities (internet research, cold calling, etc)

With these points in mind, commissions may be defensible in traditional sales environments, not because they motivate salespeople to sell, but because they motivate them to prospect!

Of course, in our case, this argument is moot because we are definitely going to free salespeople of the requirement to generate their own sales meetings.

Commissions as a hedge against non-performance

The obvious problem with the argument that performance pay provides management with a hedge against the costs associated with salespeople’s non-performance is that the same argument could be applied to everyone in the organization.

But, then sales is a special case for a couple of reasons:

1. The performance of salespeople is highly variable
2. It is easy to isolate the contribution that a salesperson makes (this would not be so easy in the case of a line worker)

As we’ll discuss shortly, SPE inverts these two reasons:

1. The output of the sales function ceases to be highly variable
2. While it’s easy to measure the capability of a salesperson, it is difficult (if not impossible) to isolate the contribution that person’s activity make to the organization as a whole

First, however, let’s consider the wider (and more terrifying) implications of performance pay.

Management abdicates

We’ve discussed earlier that you cannot manage an autonomous agent (these two concepts are antagonistic). Performance pay makes this contradiction explicit! In other words, when a significant component of a salesperson’s pay is performance based, management has formally abdicated its responsibility for sales.

In so doing, management has telegraphed to salespeople that selling is optional! It is now up to individual salespeople whether or not they generate sales – and in what quantity.

If a salesperson is capable of selling, the real cost of their non-performance is not their salary, it’s the profits that the organization does not earn when production is sitting idle!

If a salesperson is not capable of selling, the real cost of their non-performance is still not their salary, it’s the sales opportunities that are lost that could have been won if they were attended to by a more capable individual.

Variability diminished

If sales appointments are the primary driver of sales (and it’s rare that they are not), we can significantly reduce the variability of sales by:

1. Fixing the volume of sales appointments (the same number of appointments, week after week)
2. Increasing the volume of appointments (as the appointment volume increases, the variability of the entire sales function reduces)

Of course, our standard model does this by ensuring that salespeople do nothing other than sales appointments and by ensuring that each salesperson performs ten-times the volume of appointments they would perform in a typical sales environment.

Capability

Many of our silent revolutionaries report an increase in salespeople’s capability. There are three contributing factors here:

1. Predictably, when organizations reduce the size of their sales teams, they retain their more capable salespeople
2. When salespeople do nothing other than sell – four appointments a day, five days a week – they get good at it (or they rapidly conclude that sales is not the right career for them)
3. With control over salespeople – and with accurate and current data – sales management finds it easy to institute a process of ongoing improvement

Salespeople’s position on commissions

If there’s one thing I’ve learned in the last 10 years or so, it’s that sales managers are uniformly terrible at predicting their salespeople’s reaction to our standard model.

Almost without exception, sales managers predict outrage from their team members – perhaps even a mass exodus of talent! And the one component of our model sales managers predict will be the most offensive is the elimination of commissions.

In reality, salespeople’s reaction to this proposition tends to be shocking for exactly the opposite reason. It’s shocking how comfortable salespeople are to give up both their autonomy and their variable compensation plan!

The reason salespeople tend to be so compliant is very simple.

Salespeople (contrary to popular opinion) do not live in a parallel universe. They are a part of the same dysfunctional reality that is causing the rest of the organization (including management) so much pain.

Salespeople may have different theories about the source of their particular set of issues – and they may propose different initiatives as a remedy to these issues – but, when presented with the evidence, salespeople recognize (often faster than management) that a significant number of sales problems, production problems and management problems can be tracked back to the same root cause (their autonomous mode of operation).

And make no mistake; salespeople have more than their fair share of complaints:

1. They hate the volume of clerical and customer-service work that prevents them from engaging meaningfully with potential and existing customers
2. They don’t enjoy spending their evenings in hotels entering data in the CRM, generating expense reports and writing proposals
3. The resent the continual conflict over the allocation of commissions (particularly when accounts span multiple territories)
4. They hate having to advise clients that their promises will not be met – and they resent the fact that they have to live with the continuous uncertainty over production performance
5. They don’t enjoy the underlying – and constant – conflict in their relationships with production, customer service engineering, management and even finance

It may be true that salespeople are in love with the notion of the salesperson as a lone crusader, but salespeople are also realists. They quickly recognize that, on balance, the proposed environment will be infinitely more rewarding to work in.

Sure, they sacrifice their autonomy but, so what? They each get a dedicated executive assistant (sales coordinator) who will free them to do nothing but sell.

And sure, they have to transition from commission to a salary but, what of it? Salespeople understand that the dynamics of the environment in which they operate rob them of the financial upside they signed-on for. And, the truth be known, salespeople have never been entirely comfortable with the notion that they are innately lazy: prepared only to do the right thing on the promise of an incremental financial inducement.

The new compensation plan

So, it’s out with commissions and in with a new compensation plan.

But there’s not much to the new plan. The idea is simple: we pay people what they are worth (perhaps a little more).

And that’s it!

In practice, you should pay salespeople enough to ensure that compensation is no longer a regular topic of conversation – and then insist that they perform the activities required for the organization to achieve its objectives.

So here, from management’s perspective, is the fundamental difference between the two compensation plans:

With performance pay, we make optimal performance optional – and then we attempt to exert control through a compensation plan that underlines salespeople’s autonomy with every pay check!

With salaries, we take the discussion of money off the table. Salespeople willingly subordinate to a central schedule. And they perform necessary activities because they are asked to (and, because those activities are congruent with both salespeople’s job descriptions and the reasonable interests of the organization).

This new plan, then, is not even new: it’s exactly the same plan we use to compensate everyone else in the organization!

And when it comes to calculating salespeople’s salaries, there are no surprises here either. As with all employees, there are two considerations:

1. Replacement cost (how much would you have to pay for another person with a comparable set of capabilities?)
2. Asking price (how much will you have to pay the current candidate to ensure that the compensation plan is no longer a regular topic of conversation?)

It should go without saying that it would be foolish to propose that salespeople (or any team members, for that matter) take a cut in pay when you transition to this new model.

Most of our silent revolutionaries shift their salespeople to a salary that is equal to, or slightly greater than, their average total earnings (typically judged over a three-year period).

If you think about it, both parties are getting a terrific deal here.

1. Salespeople are receiving a not-insignificant pay rise. Obviously the potential to earn a figure is worth nowhere as much as the same figure, guaranteed.
2. Management is increasing the volume of effective work performed by each salesperson by ten times. To achieve the same increase in a typical sales environment, management would have to add nine more salespeople for every one they currently employ! In the new model, the cost is limited to an incremental increase in the salesperson’s compensation and the cost of a sales coordinator.

The other artificial management stimulants

This debate about commissions is like Hydra (the many-headed monster). You successfully lop-off one head and another appears.

I fear that, even if I’ve done a reasonable job of convincing you that there’s no place for commissions in the reengineered environment, your very next question might be: but, what about bonuses.

My observation is that bonus plans have a couple of problems:

1. Because the bonus is remote from the positive behaviors that drive the desired outcome, the first installment of a bonus is a pleasant surprise and subsequent installments are viewed as entitlements
2. Bonuses suggest to team members that they are responsible for outcomes when, in fact, managers should own this responsibility – accordingly, they tend to dis-empower managers

It is certainly true that some degree of variability is required where compensation is concerned. However, my position is that standard salaries provide the necessary flexibility. As your team members become more capable, their market value increases, meaning that you are obliged to grant them pay rises when (or ideally before) they request them.

I suggest that there is absolutely nothing wrong with the traditional contract between employers and employees. Employees want to be able to perform rewarding work in a secure environment. If they were really seeking uncertainty and boundless riches they would not have signed-on to be employees in the first place.

The other stimulants (targets and quotas) are problematic for the same reasons as commissions and bonuses. They tend to suggest that team members own outcomes.

In a team environment, the team cannot own the responsibility for anything! There is no collective consciousness – only a group of individuals. It is critical, therefore, that the manager owns the responsibility for the desired outcome and that team members own the responsibility only for the activities assigned to them.

Here, a military example is illuminating. Imagine, rather than allocating discrete responsibilities to each of his units a commander were simply to assemble all his troops and exhort them to take Berlin!

Reinventing management

On the subject of management, it’s important to recognize that the transition to a reengineered sales environment is extremely difficult for sales managers.

As a result of this transition, sales managers find themselves in a position where down is up and up is down. If sales managers were to refer to a list they had compiled before the transition of everything they know for sure about sales, almost every statement on that list will now be false.

Consequently, it is not sufficient to reengineer the general sales environment. You must also rebuild from scratch the sales manager’s method of operation.

* * * *

You now have a sound understanding of the theory that underpins Sales Process Engineering. Part Two of this book will show you how to convert all this theory into practice.

Let’s consider why.

Imagine you have something to sell – a widget, say. Tell me, for how much should you sell it?

The answer to that question is more obvious than you think. To calculate the optimal sale price you don’t need to know:

1. Raw material cost
2. Labor hours
3. Overhead allocation ratios

You don’t need to know any of that because the answer is: you should sell that widget for as much as you can get!

That’s right, the optimal sale price is not determined by internal factors, it’s determined solely by what the market is prepared to bear. You cannot – let me say it again, CANNOT – calculate the price of a product.

All you can do it estimate it. And you must recognize that, in most situations (certainly in those environments where we’re most likely to see T&M billing) it’s simply impossible to estimate the the price with a high degree of accuracy.

The primary assumption behind time-and-material billing (T&M) is that you can calculate the price of something by simply counting the number of minutes required to produce it.

But you can’t. Effort does not equal value. In many cases, it doesn’t even approximate it!

One of the problems with time is that it’s inherently measurable. Because it can be counted in intervals as small as a minute, we’re lulled into the false impression that we are calculating (as opposed to estimating) value.

The alternative

How then should we estimate value? And how should we track productivity?

If we use time to estimate, it’s important to recognize that we’re using time as a proxy (a yardstick) for value.  Time does not equal value.

My advice, where time is concerned, is not to estimate time in increments any smaller than half a day. In other words, do not ask, how many hours will this job take?, ask how many half days?

Where short-duration tasks are concerned, my preference is to ignore them altogether. Just factor them into the multiplier you are using for half-day slots!

I’d rather, however, that you find some other proxy for value. We’re working with a bookkeeping firm currently. When I challenged them to identify a better proxy for value than time they came up with a great one – transactions (debits and credits). It’s obvious that the volume of transactions processed is a greater determinate of client value than time consumed.

And how do we track productivity? Well, we start by burning timesheets!

You DO NOT make money out of expending effort. You make money out of delivering value. And the connection (in most cases) between these two factors is very tenuous.

Timesheets encourage team members to be busy. This drives multitasking, and the hogging of work that could otherwise be passed-off to lower-paid people. Both destroy value. Furthermore the maintenance of time-tracking systems, consumes enormous resources and the completion of timesheets insults your team members at 10-minute intervals.

You need to recognize that individual jobs (or projects) do not make you money. Your portfolio of work makes you money. Once you realize this, you can focus on managing the portfolio, rather than the jobs.

You don’t make money by keeping your team busy. You make money by delivering jobs. And the two are NOT the same thing. People work best in fits and starts. And team work necessitates relay-racer behavior (person B hovers, waiting for person A to finish his work – and then sprints to hand-off the job to person C).

You need to mobilize your team to get jobs out. Think of the pit crew in a Formula 1 team. Timesheets are not conducive to this environment.

Hi Justin

Just wanted to ask if you’ve designed a way of managing the sales budget of a company.  In other words, how to review if the salespeople are meeting the budget or not.  Most companies are affected by the end-of-the-month syndrome, and at the same time, salespeople – once they meet that month’s budget – are not interested in selling more.  They’d rather stay still and avoid cannibalizing next month’s sales.

Thanks
Alejandro

 Here’s my response:

Alejandro

Good to hear from you!

Here are the two steps we take to eliminate these problems:

Eliminate monthly budgets – in favor of a T/cu type measure (Throughput per appointment-slot-consumed).

Eliminate commissions. Pay people what they are worth and make performance compulsory.

Justin

Alejandro’s question was a reminder of just how dysfunctional most sales functions really are. His email set me thinking: why have managers (us) historically designed the sales environment this way.  (Presumably, it’s not because we’re daft, or ill-intentioned.)  Among the assumptions that underpin the design of the sales function, there must be some that are false … what are they?

Three fallacious assumptions

Here are some assumptions that I suspect will fail to emerge unscathed from an exposure to reality.  Feel free to critique my selection, or suggest your own.

Salespeople should be responsible for sales. Really? In how many organizations do salespeople actually have significant control over sales? What percentage of your transactions are the result of your salespeople:

1. Originating their own opportunities (as opposed to responding to an inbound enquiry)?
2. Prosecuting these opportunities single handed, without assistance from engineering, estimating or management?

If salespeople aren’t responsible for generating the greater majority of sales (all by themselves), wouldn’t it make more sense to hold them responsible for just the activities that they actually do perform?

Of course, in our model, we have salespeople performing only business-development meetings – nothing else.  (Marketing originates opportunities, project leaders design solutions, etc). This means that they are accountable for only what they do in these meetings.  “Sales” is management’s responsibility.  Just like “production” is the responsibility of your production manager – not a lathe operator or a claims processor.

Monthly is a sensible measurement frequency for sales. It’s true that we calculate our profitability monthly. But does it follow that all internal processes should also be measured just once every 30 days.  Consider your on-time-delivery performance, for example: should you stop and calculate this just once a month? Of course not. It should be calculated for every delivery.

We calculate profitability monthly because profitability is the consequence of a number of events that occur at different frequencies (and out of synchronization with one another). A more frequent calculation would probably generate less information, not more.

If the responsibility of the salesperson is to maximize the velocity of sales opportunities (and, yes, it should be) – and, if the salesperson’s effort is expended in discrete packets called appointments – why wouldn’t we calculate the progress of each opportunity, relative to each appointment?

As with on-time-delivery performance, you may choose to view the resulting number in the form of a rolling average, but that doesn’t alter the fact that the relevant time-horizon for the activity in question is an appointment slot, not a month.

Commissions drive positive behaviors.  I had lunch with a director of a large (publicly listed) technology company in Sydney the other day. After hearing my position on salespeople and commissions, he asked about the negative consequences of eliminating commissions (replacing them with salaries).

I answered his question – as I usually do – with my own question. I asked him if he could first detail the positive behaviors, exhibited by salespeople, that he would be comfortable to attribute to the existing commission plan.

As usual, this question (an innocent one, I’m sure you’ll agree!) was met with a silence you could cut with a knife. He admitted that there were no positive behaviors – only negative ones! And he concluded, of his own volition, that it was probably more appropriate to try and find a justification for the existence of the commission plan than it was to look for a reason to not eliminate it.

As I’ve suggested before, if sales are important for your firm, I suggest that you identify the critical behaviors that contribute to sales, and then make them compulsory (commissions signal to salespeople that these behaviors are optional).

Please comment.

Towards the end of our conversation, he asked if I thought much market intelligence could be gleaned from customer surveys.

I answered (almost instinctively), data, yes; but, intelligence, no.

When pressured for a more coherent answer, I explained that I had never seen customer survey responses that provide any market intelligence. Never. Never. Never!

My position is that, to qualify as intelligence, survey results must be:

1. Unexpected (if you know the answer already, why ask the question?)
2. Actionable (if data doesn’t cause you to do something differently then where’s the benefit?)

Think about the survey results you’ve seen.  How often have you been surprised?  And how often have you been motivated to take some specific action?

Now, of course, I’m not advocating ignorance. Market intelligence is critical. You shouldn’t try to survive without it. It’s just that surveys are a pretty poor way of generating intelligence.

Here’s the thing …

In your business, you bank behavior, not attitudes.

My advice then, is stop measuring attitudes and, instead, measure behavior. Run controlled experiments. Change offers, change headlines, change opportunity-management workflows (not all at once, of course) and measure the impact of changes on, enquiries, conversions and, ultimately, sales.

Now, those who make their living from surveys will argue that attitude is an antecedent to behavior. I’m sure that in most cases it is. And in some, it definitely isn’t (how many smokers genuinely believe that their’s is an intelligent choice?)

But even if this claim is true, this position contains another assumption: that surveys actually measure attitudes. Do they?

Before you answer that question, consider another: which do you think is the more accurate predictor of a client’s future behavior:

1. Their current attitude
2. Their prior behavior

Tell me if you disagree but, to my mind, it’s a no contest!

And behavior is much sexier than attitude for another reason too: it’s inherently measureable (in a much more objective sense than the surveyor’s 1-5 scale).

If you value market intelligence, can the surveys. Objective management is a philosophy, not a once-a-year event!

When you are mapping your workflow, how do you determine the ideal level of granularity?  In other words, how much detail is too much?

Conventional wisdom is that:

1. Planning and execution are two discrete activities (remember this assumption)
2. The key to good execution is lots and lots of planning (which tends to mean planning at a very granular level)

Conventional wisdom is wrong, however.  And this is evidenced by the many examples of poor execution we encounter on a day-to-day basis.

My argument (and this is classic TOC) is that, above a very low detail threshold, more planning actually means worse execution!

In other words, in most cases, to execute better, you should plan only at a high (non-granular) level.

The relationship between granularity and confidence

Let’s imagine that we are planning a simple workflow – getting to work in the morning.  Here’s a high-level (low-detail) plan:

1. Get out of bed
2. Shower
3. Get dressed
4. Drive to work
5. Log-in to computer

And here’s the low-level (high-detail) one:

1. Upon awakening, check to see if alarm is sounding
2. If so, deactivate alarm
3. If not, check to see if current time is greater than [alarm time minus 10 minutes]
4. If so, cancel alarm and get out of bed
5. If not, go back to sleep
6. And so on …

Imagine that you completed the second (low-level) plan, and then compared it with the first – and answered this question for each plan?

Rate your degree of confidence that reality will play-out as specified in the plan? In other words, how confident are you that each of these activities will be performed in exactly the sequence specified?

My guess is that you will be highly confident where the first plan is concerned – and not-at-all confident in the case of the second.

In fact, it’s probably worse than that. In the case of the second plan, you will likely be certain that reality will not play out as specified by the plan.

The irony is that the second plan creates a perception of accuracy, in spite of the fact that your confidence in the plan is practically zero!

Finding the inflection point (essential and non-commutative)

It’s tempting to assume that there’s a linear relationship between granularity and confidence (that they vary in direct proportion).  And, if that’s the case, there’s no correct answer to the critical how much detail question.

The reality is that the relationship is definitely non-linear: as the granularity of the plan increases incrementally, beyond a critical detail threshold, your confidence in the plan plunges rapidly towards zero.

Here’s why:

1. All the activities in the first plan are essential (you simply can’t skip any)
2. Each activity (relative to its predecessor or successor) is non-commutative – meaning that you can’t swap the sequence without dramatically changing the outcome

However, with a slight increase in granularity, these two conditions cease to apply: not all activities are essential and the outcome can conceivably be delivered with an alternative sequence of activities.

So, my suggestion is that you should plan only to the level of detail where you are confident that all activities are (a) essential and (b) non-commutative (relative to their neighbors).

Semantics are important!

I can almost feel your incredulity at this point! How can a plan at this high level of detail be sufficient for execution?

And your reaction is warranted … because it can’t be!

The high-level plan that I’m recommending is sufficient for planning – but not sufficient for execution.  The thing is that planning must continue after the creation of the plan!

To clear the confusion we need two words for planning:

1. Pre-plan planning (the planning phase, terminating with THE PLAN) – let’s call this planning
2. Post-plan planning (the execution phase) – let’s call this fine-tuning

So, in summary, here’s the solution:

1. Create a plan that contains only the degree of detail that enables you to be highly confident that reality will play-out as prescribed
2. Accept that the level of detail in the plan is insufficient for execution
3. Manage prioritization decisions on a day-to-day basis to ensure that reality tracks to the plan

In the post referenced above (quick-and-dirty approach), I talked about the importance of centralizing scheduling and building a management information system (MIS).

If you re-read that article it should be clear now that both of these initiatives are essential for the management of execution.  It should also be clear that the MIS should be designed to enable management to fine-tune the prioritization of low-level tasks so as to ensure the compliance of the undertaking with the high-level plan.

(Execution management is outside the scope of this posting. To explore the subject further, read my brief overview of Critical Chain Project Management.)

The work we do in these non-traditional (for us) environments is very gratifying and, often times, generates enormous payback for our clients.

So that just started me thinking.  Can I generalize from our approach to the sales environment — and from the work we do in other environments — to come up with a general approach to process improvement?  My plan is not to write the book on process engineering (others have done that); rather, it’s to provide a step-by-step method that you can implement with your notebook computer in one hand, a coffee in the other, and look of gleeful masterly in your eye!

Here’s that method; in five simple steps.

1. Define the goal and necessary conditions

What is the goal of this environment?  And what are the conditions that must be met in order for your achievement of that goal to be valid?

Sadly, answering these questions does not really power you forward.  However, they must be answered because the answers provide the context required to answer all the other questions you’ll encounter as we work through this process.

So, if you’re building a customer service team, your goal might be to: resolve customer issues rapidly.

Necessary conditions might include: the maintenance of sufficient protective capacity to ensure resolution times are within customer tolerances and the avoidance of unnecessary expenditure.

Nothing too fancy here!  Just the basics.

2. Map it

This is the fun bit.

You’ll need a copy of Microsoft Visio (Google: Visio free trial) or a sheet of graph paper and writing implements.

The trick here is knowing what to chart.  You want to map a high-level workflow only.  You’ll know if it’s high-level because:

1. You’ll only need three symbols (an activity box, a state (input/output) circle and a connector (line)
2. The workflow will be linear: no loops or recursions
3. Everyone involved will complain that it’s too simple

Here’s an example of what the finished result will look like.

[click image to view full size]

2. Division of labor

You saw this one coming!

As part of creating the workflow, you should consider division of labor (who does what?).  You can see in the image above, that each role as been assigned a swim-lane.

Generally speaking, you’ll multiply productivity by:

1. Eliminating multitasking
2. Having specialists perform work that corresponds with their expertise

However, you need to balance the theoretical productivity improvements with some practical concerns:

1. The possible cost of additional personnel
2. The multiplication in process complexity as the number of hand-offs increases

Here are the general guidelines I use:

1. Split activities that are obviously dissimilar (technical drawings and booking flights, for example)
2. Split activities that are performed in the field from those that are performed in the office
3. Split pure telephone work from clerical work (for example, rather than having clerks chase the pre-requisites required for their work, have a dedicated phone person who manages pre-requisites)
4. Split work that requires high-level decision-making from pure-vanilla work

More importantly, you should consider which resource in your team is likely to become the constraint (bottleneck) as volume increases.  You should focus your attention here first.  For example, if you are processing mortgage applications and you have a person with specialist knowledge who has the potential to be a bottleneck, consider routing all plain-vanilla work (and phone work, etc) around this person.

Also, bear in mind that some people are specialist generalists.   In other words, they add value by being across a lot of stuff!  Managers fit into this category — as do customer-service representatives.  So be careful not to kill the goose that lays the golden egg.

4. Centralize scheduling

It’s here that you earn your keep!  If there’s one defining characteristic of EVERY process we build at Ballistix, it’s that we always have a dedicated process coordinator.  A person whose sole responsibility is to coordinate the workflow — but who is not actively involved in any of the core activities within the workflow.

I would go so far as to suggest that you don’t have a process until you have a person who’s sole responsibility is to keep the wheels turning day after day.

In our traditional (field) sales environments, this person is the sales coordinator.  A person who sits behind each salesperson and who operates essentially as that person’s Executive Assistant.  In an inside environment (customer service, inside sales, pre-production, etc), we’re more likely to have one person coordinating a larger team.  In many cases, this person will also take responsibility for pre-requisite management (see earlier discussion).

Here are the activities a coordinator would typically perform:

1. Chair daily (or twice-daily) work-in-process (WIP) meetings
2. Determine case (job) and task priorities
3. Overview the general operation (flow) of the environment
4. Ensure data integrity (accuracy and currency) in your CRM and/or ERP
5. Manage management-information system (see next section)

In most cases we recruit young, smart, sassy graduates to fill coordinator roles.  Generally, you don’t want an engineer in this role.  In fact, it’s better to retain an engineer to design the environment so as to ensure that you don’t need an engineer to coordinate it!  Of course the design of a robust environment is exactly the subject of this article.

5. Management-information system (MIS)

You know what they say: if you can’t measure it, you can’t manage it.

Well, I wish they wouldn’t say that.  This maxim is often interpreted as an exhortation to measure everything.  And if you measure everything, you’re measuring nothing.

What I mean is that no data is bad, but too much data is just as bad.  What we need is information.

Eli Goldratt defines information as the answer to the question asked … and I like that!  It means that rather than starting with measurements, we should start with questions.

Here are some questions that I think should be answered in every environment:

1. When should I release work to the team (if you release work as it appears, the team will certainly multitask itself into a total logjam within days!)?
2. What completion date should I quote for this case (and be confident that it will be delivered on time)?
3. When must I expedite a case to ensure its on-time completion?
4. When should I avoid the temptation to tamper and leave well alone?
5. What is the load on all team members?
6. When the flow is disrupted, what is the most common cause of that disruption?

Now, obviously, this list is not exhaustive but these answers are, to my mind, the critical ones.  So, do you think the standard reports in your ERP or CRM are designed to answer these questions for you?  Of course not!  Those reports are designed to satiate the noisy measure everything brigade.

So, you need to build your own MIS.  Two ways to do it:

1. A physical, production-style buffer board
2. A fancy, electronic business-intelligence system that generates real-time reports from the data sitting in your CRM/ERP’s data-store

Let’s talk about the first option.

You want a board that represents time and capacity (we’ll just work with two dimensions for now!)  We’ll track time on the horizontal access with TODAY on the far left and X-DAYS on the far right (where x is the maximum lead-time of a healthy case).  So time moves from right to left — just like you’re on a train, looking out the window. The vertical axis is designed to limit the number of cases that can occupy any one time-slot.

The coordinator creates a magnetic card for each job.  Each card contains the list of tasks within the workflow.  Cards also have stickers attached — where each sticker indicates a pre-requisite.  At regular intervals the coordinator moves all jobs one time-slot to the left — to indicate the passage of time.  Additionally, stickers are removed as pre-requisites are collected and tasks are checked as they are completed.

We would generally complement the buffer board with an Excel workbook that auto-generates the cards — and that enables the coordinator to track the completion dates of key milestones so that we retain a rudimentary history of jobs once they leave the board.

Here’s an example of a client’s buffer board.  Pretty? No.  Effective? Definitely!

[click image to view full size]

So how about the second option?, I can hear you asking.  Can’t it be electronic?  We so like electronic.

Well, yes it can and, in some environments (high volume, high complexity) it has to be.  But electronic costs real money.  And another disadvantage of electronic is that it tends to be less visible — less accessible.  We resolve this in some environments by displaying key metrics on plasma screens … but now we’re talking even more money!

To create an electronic MIS, you build the reports in Excel (our drug of choice), Crystal Reports, Cognos or similar.  You then integrate the reports with the data from your CRM or ERP (or often both).

Of course, with electronic, if you can dream it, you can do it … but this tends to create as many problems as it solves.

Following are some samples of some of the prettier reports that our clients’ MIS’s contain.

[click images to view full size] 

The best advice I can give you with respect to the design of the MIS, I’ve left until last.

Don’t start by designing the reports, start by designing the WIP meeting.  That’s right, determine the questions that you must answer, in what sequence to have a fast, productive, daily (or twice-daily) WIP meeting and then design your MIS to deliver the necessary data in exactly that sequence.

These five steps cover a lot of ground, but I think they’ll provide a valuable framework for your next process-improvement initiative.  If you’d like me to fill in some gaps, use the comments link at the start of this article to ask questions.

Philosophies collide!

I was presenting a workshop in Darwin (Australia) recently when the manager of a (government-funded) organization objected that our methods focus only on the financial bottom line — and that they are not conducive with the concept of triple-bottom-line management.

Now, I must have had my head in the sand for some time because, up until then, I had never heard of this peculiar concept.

The idea is that an organization shouldn’t strive to improve just one bottom line (the financial one), rather it should also pursue the improvement of two (equally important) additional bottom lines: an environmental one and a social one.

Now, thinking readers will have already recognized two obvious objections to the triple bottom line:

1. Division of labor (As Adam Smith pointed out, the economy as a whole benefits most when organizations specialize).
2. Practicality (there can be no single standard of ‘environmental good’ or ‘social good’; only multiple measures that will ultimately conflict with one another)

But the argument that I presented in Darwin was simpler — and I suspect stronger — than either of these.  And it it was pure TOC!

In TOC terms, the concept of the triple bottom line amounts to an argument that the organization should pursue multiple goals.

Is this possible?

Well, we know that a business is a system (a goal-seeking system).  This means that a business is not a collection of independent parts — all the parts are connected (directly or indirectly) to one another (a part that wasn’t connected one way or another would not be part of the system).

We know that each of the parts of the (business) system has finite capacity: accordingly, the system has a whole has finite capacity.  (If this wasn’t the case, it would have achieved it’s goal in the instance in which it came into existence!)

So, with these facts in mind, let’s think what happens when a business attempts to pursue multiple goals simultaneously.  Inevitably, the pursuit of each goal requires a different allocation of resources from that finite resource pool.  (This is called ‘resource contention’.)

The only way to resolve resource contention is to rank the three goals in order of priority.

The only way to perform this ranking is to nominate one of these goals as The Goal, and the others as Necessary Conditions.  There is simply no other alternative in a finite-capacity system.

For this reason, I propose that a business cannot pursue multiple goals and that the notion of a multiple-bottom-line is discredited!

Post scripts

In the TOC world, the following definitions apply (my wording):

The Goal: the output the system is designed to produce (measured in units called Throughput).

Necessary Conditions: conditions that must be complied with in order for the achievement of the goal to be valid. 

Necessary conditions are bi-modal (you either comply or you don’t). If we’re dealing with a typical business, the goal is to maximize the return on shareholder funds.  An example of a necessary condition is ‘cashflow’.  Maintaining adequate cash-on-hand is critical.  More cash than that which is adequate provides no additional benefit.

Sadly, the workshop delegate was not happy with my answer.  He left (in a huff) in the lunch break.  By the way, he was correct to object that our method is not conducive with triple-bottom-line management.  It isn’t.  But neither is reality!

The essence of my argument was that, in environments where transactions are small in frequency, but large in magnitude ($’s), the traditional approach to forecasting destroys information — rather than creating it.

The traditional approach: statistics

Here’s an example.

Let’s assume that you are the manager of a project environment in a technology company.  Your job is to deploy teams of technical experts to deliver large projects that have been sold by your sales department.

Your challenge is that you need to make resource-allocation decisions based upon both current and future orders.

You ask your sales department for an estimate of future sales and — after much gnashing of teeth and grinding of gears — you are provided with a forecast.  This forecast advises that sales next month will be $653k; the following month: $432k; and so on.

These numbers seem strangely precise, considering that they are estimates.  Furthermore, the forecast doesn’t give you any indication of how confident you should be of these estimates falling within a particular range of values.

Suspicious; you ask the head of sales how these numbers were compiled.  He advises that the following method was used:

1. Salespeople used a set of rules to assign a probability to each opportunity upon which they are working (these rules prompt salespeople to consider factors like number of competitors, seniority of decision-makers, and so on).
2. Salespeople’s data was aggregated so as to provide a risk-adjusted estimate of future orders by calendar month.

Ask yourself, how useful is this forecast?

The forecast tells you that orders in two months’ time will total $297k.  When you enquire into this number you discover that it primarily consists of a $635k order that Sales believe they have a 34% probability of winning.

What decisions can you make on the basis of this estimate?  Can you set-aside 34% of the necessary resources?  Should you recruit new personnel?  And, in either case, can you pay for these resources using the $216k revenue that constitutes the risk-adjusted value of the $635k sale ($635k x 34%)?

Of course not.

If Sales fails to win that order in two months’ time as projected, the project will place no demand whatsoever on your resource pool, and will generate precisely $0 in revenues!

To use statistics in this environment is obviously foolish because orders are won so infrequently.  The problem is compounded by assigning probabilities to opportunities based primarily upon salespeople’s (subjective) opinions.

So, if this approach is so obviously foolish, why do so many organisations adopt it?

I can think of a handful of reasons:

1. Because it does make sense in environments where transactions are high in frequency and low in magnitude
2. Because it appears (at first glance) to be scientific
3. Because senior management (or other departments) demand it
4. Because the CRM came with this capability built-in
5. Because no one is aware of an alternative approach

An alternate approach: heuristics

Let me propose an alternative approach — the approach we use internally.

But first, let’s remind ourselves that, because the dataset upon which we are reporting is not statistically significant, our approach must not make use of statistics.

The alternative approach is remarkably simple.

Sales maintains two sets of assumptions for each opportunity:

1. Optimistic (but not exuberant)
2. Pessimistic (but not paranoid)

So, for each of the critical opportunity variables (opportunity value and commencement month) both an optimistic and a pessimistic value is recorded — and revised weekly.  These values are not calculated, they are arrived at by discussion in sales meetings.  In other words, a value for each variable is negotiated by the salesperson, the sales coordinator and the sales manager.

Then, each week, two scenarios are created for distribution to senior management (and other departments) — the optimistic and the pessimistic scenario.  Each scenario details specific projects (and the full value of each) — as opposed to risk-adjusted aggregates. Ideally, each scenario will also detail resourcing and cashflow implications.

Each department then uses these two scenarios to plan as best they can.

What happened to precision?

At first glance the second (heuristic) approach lacks the precision of the traditional approach to forecasting (statistical).

However, upon closer examination it becomes obvious that the precision of the traditional approach is an illusion.  Contrary to popular belief, the mathematical manipulation of salespeople’s opinions does not convert subjective data into objective information.

The heuristic approach acknowledges the existence of uncertainty, rather than disguising it.  Accordingly, it is significantly more useful to those who intend to use the reports provided by sales to make actual business decisions.

When you open your eyes, your bedroom appears roughly as it did the night before. Your bed is below the open window, and your dresser is still adjacent to the door.

However, a second look reveals that the curtains that normally hang above your windows are missing. As is the painting that normally hangs, slightly crooked, on the wall facing your bed.

What’s more, you notice some strange additions to your bedroom furnishings. A plaster bust now dominates your dresser. And an empty hat rack leans precariously against your bed head.

As you rise and navigate your way around your bedroom, you discover that your memory provides the accurate version of reality. In spite of the information tendered by your eyes, your curtains still hang above your open window, and your painting still adorns the wall adjacent to your bed. Your sense of touch confirms that there is no plaster bust on your dresser, and that the hat rack is also a mirage.

You only have to spend a day in a typical marketing or sales department to discover that this scenario is analogous to the environment in which management operates.

Various reports and performance indicators (both formal and informal) provide glimpses of reality. However, this feedback falls short of delivering the accurate and complete viewthat managers need in order to be truly effective.

When hobbled with an incomplete and unreliable view of reality, managers’ activities are, at best, inefficient and risk-adverse. At worst, managers unknowingly engage in activities that are harmful to the organisation as a whole.

Hence the need for an objective management structure!

Reality

By definition, reality must be the starting point for an objective management structure.

Without the ability to accurately perceive reality, measurement and, accordingly, management is impossible.

Of course, the concept of measurement presupposes something to measure.

This means that the ability to perceive reality is not enough. We need also to determine upon what aspect of reality we should focus.

It is perhaps self-evident that our focus must be determined by our goal. After all, a measurement out of context with a goal is just a number (data, not information).

If your goal were to drive from Sydney to Darwin, your current location, speed and direction of travel are information; the weather in Perth is data.

For this reason, the most critical step in designing an objective management structure is defining your goal.

I’ve had many managers assure me that this is also self-evident:

Marketing managers tell me that it’s their goal to generate sales opportunities.

Sales managers advise me that their goal is to generate sales.

In each instance, managers forget that the processes they manage are part of a larger system. As a result, they fail to recognise that their goal must be subordinated to the goal of the system.

One system, one goal!

For simplicity, let’s define a system as a set of interdependent processes.

A fine example of a system — one with which we’re all familiar — is an internal combustion engine (pictured below).

As you no doubt know from your experience with this particular system, the goal of the internal combustion engine is to generate torque (rotational force).

As depicted above, this system consists of four processes:

1. INTAKE (a mixture of fuel and air is drawn into the cylinder).
2. COMPRESSION (this mixture is compressed by the rising piston).
3. COMBUSTION (the fuel/air mix is ignited by the sparkplug).
4. EXHAUST (the resulting exhaust gasses are expelled from the cylinder).

Now, consider this question for me:

If you were the sparkplug in this system, what would be your goal?

1. Would it simply be to generate sparks (as many as possible)?
2. Or would it be to generate (a set quota of, say) 10 sparks a second?

Of course, neither answer is correct.

Your goal could only be expressed in terms that recognise the relationship between your activities and the goal of the system.

Accordingly, your goal would be something like the following:

To produce a spark at the top of each compression stroke.*

* Technically, the sparkplug produces a spark fractionally before the top of the compression stroke.

If your goal must be subordinated to the goal of the system, it’s essential for us to identify your organisation’s goal.

Why does this organisation exist?

Whenever I ask this question of a seminar audience, I get a range of answers:

1. Some managers claim that their organisations exist to manufacture widgets.
2. Others explain that their organisational goal is produce happy customers.
3. Others try and cover all bases with statements that reference the organisation¡¯s relationship with all stakeholders (shareholders, staff, the community, the goldfish in the corporate pond etc.)

In each instance, managers are confusing the goal of their organisations with necessary conditions.

The goal of any commercial organisation is, by definition, to make money.* Necessary conditions are the conditions that must be present to enable this goal to be achieved.

* Goldratt (whose work has had a significant impact on the thinking behind this article) explains that the goal of a business is to make money, now and in the future. I would argue (as, in fact, does he) that the latter part of this statement is redundant (we’re all accustomed to recognising the value of future revenues). To learn more of Goldratt’s work, begin by reading The Goal.

While the idea that the goal of a commercial organisation is to make money may initially cause managers some discomfort (believe me, it does), it also provides the clarity we need to start to piece together our objective management structure.

Unravelling organisational complexity

As our internal combustion example illustrated, all goals at a process level must reflect the contribution that the process makes to the goal of the system as a whole.

Unfortunately, organisations tend not to be as simple as our little engine.

In fact, one of the greatest challenges faced by managers is the requirement to understand the interaction between multiple organisational processes.

As has been explained in recent editions of AdVerb, this challenge has been greatly simplified by Goldratt’s Theory of Constraints (TOC).

In short, TOC recognises that the output of any system is determined by the system’s constraint (or bottleneck). It also points out that:

1. Every system has a constraint (if it didn’t, output would be infinite).
2. At any one point in time, every system has only one constraint.
3. A stable system is one where the constraint remains in one location.
4. Because the constraint is the sole determinate of system output, every management decision should reference the constraint.
5. The investment of resources in any non-constrained process or activity will produce absolutely no return (contrary to the assumptions that underpin cost accounting).

An understanding of TOC enables a manager to ignore organisational complexity when making decisions, and simply consider the impact that the options under consideration will have on the system’s constraint.

Now, because the goal of the organisation is to make money, when we’re discussing system output, what we’re really talking about is money. And when we’re discussing the system’s constraint, our consideration is the flow of money at the constraint.

TOC practitioners use a simple formula to express this concept:

Flow of Money = Throughput / Available Constraint Units

Throughput (T) refers to revenue minus totally variable costs (true gross profit).

Available Constraint Units (ACU) refers to the number of units of constraint that are available over the period of consideration. (If the constraint is a machine, the Constraint Unit is likely to be time-based. If the constraint is a salesperson, the Constraint Unit will be an appointment slot.)

As we’ve established, the goal of every process within our organisational system should reference the organisational goal: maximising the flow of money at the system’s constraint (or T/ACU).

Let’s pause now, for a practical example.

A two-process organisational system

Consider the simple organisation pictured below.

If we assume that the constraint in the organisation above is in the production process, what is the goal of the sales process?

Is it to produce as many sales as possible?

Obviously not!

The sales process should produce enough sales to ensure that the production process operates at peak capacity, all the time. To produce more sales than this would waste resources (and annoy the marketplace).

But, that’s only the half of it.

If the production process is constrained, the sales process should also produce the kind of sales that are going to deliver the greatest return on the Constraint Units consumed.

Here’s the solution:

System Goal: maximise T/ACU (measured at point of sale).

Production Process Goal: maximise T/CU (measured at point of sale).

Sales Process Goal: maintain a production buffer of optimal size (x days’ worth of unstarted work in progress). Maximise the value of the production buffer (T/CU).

To translate this into plain English:

The goal of our simple organisation is to maximise the yield (Throughput) it gets on its Available Constraint Units. (Because the output of a system is determined by the system’s constraint, this is equivalent to saying that the goal of the organisation is to make money.)

The goal of the production process is to maximise the yield it earns on every Constraint Unit it consumes.

The goal of the sales process is to:

• Maintain a buffer (or inventory) of unstarted work in progress (or orders) of an optimal size. If this production buffer is too small, there is a risk that the production process may sit idle from time to time. If it is too large, this indicates that the sales process is wasting resources that would be better deployed at the constraint.
• Maximise the value of the production buffer. The value of the buffer is the total Throughput that the buffer represents, divided by the number of Constraint Units that will be consumed to realise this Throughput. The inference here is that not all Throughput dollars are equal in this scenario. The sales process must focus on winning the sales that produce the greatest yield on the scarce Constraint Units.

The sales process dissected

We have now accumulated all the theory we need to apply our focus to the sales process.

I’m going to assume for simplicity’s sake (and because it is often the reality) that your system constraint is your sales process. In other words, I’m assuming that your production process can handle all the sales that your sales process produces (at least for the duration of this discussion).

Because we’re now shifting our focus from the system (which is a collection of processes), to one particular process within that system, we’re now going to have to think at a more granular level.

Our interest now shifts from determining which process is constrained to determining the exact location of the constraint within the constrained process.

Let’s consider a simple sales process (designed around Relationship-centric guidelines), containing three personnel:

The basic responsibilities of each person are:

1. Marketing Coordinator. Producing sales opportunities.
2. Sales Coordinator. Managing the salesperson’s diary.
3. Salesperson. Negotiating transactions.

Before we even attempt to manage this process, we need to determine exactly where the constraint is.

Now, the nice thing about operating at this level of granularity is that you get to choose! At this level, it’s quite easy for you to shift resources around to ensure that the constraint is exactly where you want it to be (and to ensure that it stays there).

If you consider that, in any process, by definition, it is only the constraint that operates at 100% capacity, ask yourself, who in the process above would you least like to be idle?

It’s obvious, isn’t it?

You would like your salesperson to be operating at 100% capacity, all the time. This is because your sales process will generate a greater flow of Throughput when your salesperson is fully utilised than it will when either your sales or marketing coordinator is flat-out (and your salesperson has idle time).

If you want to ensure that your salesperson is the constraint (and stays that way), you simply over-resource the marketing coordinator and the sales coordinator. In practical terms, this means ensuring that they never have to operate at full capacity (at least where their critical tasks are concerned).

As we did in our previous example, we’re going to take one more precaution to ensure that the constraint stays put.

We’re going to build an inventory (or buffer) of unallocated sales opportunities to ensure that the salesperson never has to sit idle.

The resulting process is pictured below:

Now, let’s apply the same logic to this sales process that we applied a moment ago to our simple organisational system.

We should first remind ourselves that the sales process pictured above is part of a larger system ¡ª the business as a whole.

We already know that the goal of that system is to make money.

We also know that the money that the system makes is determined by the organisational constraint, which happens to be the salesperson.

The salesperson’s Constraint Unit is an appointment slot.

 Accordingly, the goal of the system as a whole is to maximise T/ACU (or Throughput / Available Appointment Slot).

Now, if we shift our focus to the sales process, it is obvious that the goal of the sales process must be the same as the goal of the system. (The sales process is the constrained process.)

Let’s look now at each of the personnel operating within the sales process.

The salesperson’s goal should be to maximise his Throughput per appointment slot consumed (or Throughput / Appointment). It’s worth noting that the salesperson’s goal is not simply to maximise Throughput, nor is it to maximise his conversion rate. Each of these goals is likely to result in behaviour that is sub-optimal in the context of the system as a whole.

The sales coordinator’s goal is to keep the salesperson fully utilised. In other words, it is her responsibility to ensure that every available appointment slot is filled. (An empty appointment slot represents an opportunity cost equivalent to the average Throughput / Appointment).

As you’ve no doubt guessed, the marketing coordinator is responsible for both the size and the quality of the opportunity buffer. The buffer should be maintained at its optimal size (measured in days¡¯ worth of appointments). It should also be composed of opportunities that are likely to yield the highest Throughput / Appointment.

Now that we have established an objective (and systemically congruent) goal for each person in our sales process, performance indicators are pretty much self evident:

Function	Goal	Performance Indicator
Sales Process	Maximise yield on Salesperson	Throughput / Available Appointment Slot
Salesperson	Maximise yield on Appointments	Throughput / Appointment
Sales Coordinator	Ensure salesperson is fully utilised	% of Optimal Utilisation (Optimal Utilisation is 100%)
Marketing Coordinator	Maintain opportunity buffer at optimal size. Maximise value of opportunity buffer	% of Optimal Days. {Throughput / Appointment Slots x Probability}*

* means sum of. Throughput, Appointment Slots and Probability are all estimates. Probability refers to the probability of your converting that opportunity.

Management just got easier

Well there it is: an objective sales process management structure!

Each person’s goal (and the accompanying performance indicator) is congruent with the goal of the system as a whole ¡ª and, accordingly, is reality based.

Each performance indicator is quantitative, rather than qualitative, meaning that no subjective interpretation of results is required.

Each person has only one or two performance indicators (one is optimal), meaning that there is no confusion as to what she should be doing, and how she will be judged.

Each person is likely to have a clear understanding of the contribution she makes to the success (or otherwise) of the system as a whole. As well as being good for morale, this discourages the development of political factions.

Creating a productive environment

As I’m sure you’ve already realised, an objective management structure will make an enormous contribution to organisational productivity.

However, there are three more initiatives we recommend you implement in order to create a truly productive environment.

1. Use run charts. It’s one thing to allocate a performance indicator to a team member. It’s another to find a way to use this information to positively influence her behaviour. If you remind yourself that a performance indicator is just feedback from the system, the solution is obvious: plot it on a graph. Better still, have each team member plot her performance indicator on a run chart. An example of a run chart follows. This was generated within Excel, however, traditional plotting paper does as good a (and some would argue, better) job.
2. Eliminate budgets and bonuses. Managers traditionally have attempted to motivate their teams with a cocktail of artificial stimulants. Budgets, bonuses, performance reviews, quotas and management exhortations are tactics invented by management to cope with the lack of an objective management structure. You’ll find that, if you simply pay your team members what they’re worth — and have each discuss his or her run chart at a weekly management meeting ¡ª your requirement for these artificial stimulants will rapidly dissipate.
3. Replace the word maximum with optimum. In this discussion I have taken a shortcut (for the sake of simplicity) and allowed myself to use the word maximum on a couple of occasions. I’ve referred to maximising the yield on the constraint and to maximising the value of the opportunity buffer.

My preference is, in both cases that you replace the word maximise with optimise.

We’ve already discovered that every process (and system) is constrained. What this means is that, if you scale a process, you will eventually reach a point of diminishing returns.

In the case of maximising the yield on the constraint, you may find the constraint moves (which is generally undesirable), or that your ability to grow your business is constrained (it takes time to recruit and train new staff).

In the case of maximising the value of your opportunity buffer, you’re likely to find that, past a certain point, increased promotional costs are likely to overwhelm any increase in opportunity value.

Now, in both instances, it will take some experimentation to determine those optimas. However, in the short term, I suggest you aim only for incremental improvements in these (and similar) situations. From a systemic perspective, it is far healthier to aim first for stability, and second for incremental improvement, than it is to set stretch goals. (The latter approach is guaranteed to wreak havoc elsewhere in the system.)

My thanks to James Powell (Viago), who provided technical assistance for this article. Viago assists organisations to apply TOC to production processes. www.viago.com.au

Over lunch, a CEO recently admitted to me that his financial controller was using his organisation’s profits to build quite a substantial commercial property portfolio.

When I asked if this was best use of his organisation’s free cashflow, he smiled, "How did I know you’d ask that question?

"The fact is," he continued, "the availability of capital is not currently a constraint on our growth."

"It might not be a constraint on your organic growth," I argued, "but I still think that money could be better invested in new client acquisition."

We spent the rest of that lunch scribbling on napkins. Together, we discovered that a dollar invested in promotional expenditure was actually providing this CEO’s organisation with a better-than 900% return on investment!

By the time coffee was served, we’d agreed that his organisation’s sales process could provide a much more favourable return on capital than even the bluest of blue-chip commercial property investments.

In our experience, this holds true for most organisations. Unfortunately, most are reluctant to invest in their sales processes because (unlike other business processes) it’s impossible to calculate a return on investment.

Science versus art

If you’re a follower of our Relationship-centric Marketing methodology, you’ll know that a sales process has inputs and outputs — just like any other business process.

You’ll know that a sales process (as the word process implies) consists of a sequence of simple, interrelated steps — just like any other business process.

And you’ll know, at least in theory, that each step in a sales process can be measured, managed and optimised — just like any other business process.

This article explains the science (and more importantly, the mathematics) behind sales process management. It will show you how to take control of your sales process and use it to fast-track the growth of your organisation.

If you didn’t pay much attention to mathematics at school, you may find this article tough-going at times. But please be sure to persevere.

I’m sure you’ll discover that your sales process is harbouring significant growth potential!

Management by numbers

If you think about it, the word management presupposes measurement. The fact is, if you can’t measure it, you simply can’t manage it.

So, to manage a sales process (or any process for that matter), we need to know what to measure. Generally speaking, we will measure inputs, outputs and time. Specifically, we’ll measure:

• Throughput (output/time)
• Productivity (output/inputs)

We’ll measure these key performance indicators (KPIs) for the process as a whole, and then we’ll break the sales process into its key components (sub-processes) and devise a set of KPIs for each component.

It’s worth remembering that your sales process is actually a component of a much larger system: your entire business. In the context of your business as a whole, revenue is a measure of throughput, and gross profit is a measure of productivity. The problem with these indicators is that they are trailing indicators: that is, they tell you more about what you have done in the past than they do about what you should do in the future.

Because your sales process is the first step in your entire organisational process, the information you collect from monitoring these performance indicators can be used to enable real-time process optimisation.

As we dissect and analyse the sales process, we’ll make references to a fictitious company we’ll call Correlex. Correlex is an engineering firm that consults to property developers. Correlex’s clients all pay a retainer of $450 a month to access its consulting services. References to Correlex will appear in indented sections, with a green sidebar, just like this one.

Measuring the process as a whole

The objective of your sales process is obviously to generate sales.

This process must be designed and managed to ensure that it delivers a sufficient volume of sales in exchange for a commercially realistic investment.

For simplicity’s sake, we’ll assume that the objective of our sales process is to acquire new clients. But don’t worry, we will be sure to take follow-on sales (and even referrals) into account.

Accordingly, the two global indicators in which we’re most interested are:

• Client acquisition rate (new clients per month).
• Client acquisition cost (cost per new client).

When we’re measuring client acquisition cost, we’re dividing the amount we invested in the acquisition of relationships by the number of clients acquired as a result of that expenditure.

When we calculate this figure, we only take into account the variable costs associated with the promotional campaign that acquired each particular client relationship. With a promotional campaign, variable costs are typically media costs. We do not factor in the fixed costs associated with that promotional campaign (the cost of creating the campaign). Nor do we include the fixed costs associated with the rest of the sales process (e.g. the cost of managing the relationship with the potential client).

We ignore fixed costs because these are the cost of operating your sales process, rather than process inputs.

Our client acquisition cost is most useful for monitoring the performance of our relationship acquisition campaigns. It’s important to remember that, unless you have a very short sales cycle, client acquisition cost tends to be longer-term performance indicator. (The term sales cycle refers to the average time span between the acquisition of a relationship and the consummation of a sale.) We’ll uncover a short-term indicator when we examine the relationship acquisition step of the sales process.

When the CEO of Correlex reviews his organisation’s sales process, he discovers that, averaged over the last 12 months, Correlex acquired two new clients each month.

To determine his average client acquisition cost, he divides his total variable promotional costs by the number of clients he acquired over this period.

In the last 12 months, Correlex had advertised in the Financial Review, and in a number of specialist publications. It had also run 4 direct mail campaigns. Accordingly, its variable promotional costs consisted of the cost of media for the advertising campaigns, and the cost of mail processing for the direct mail campaigns.

In total, Correlex invested $9,600 in order to acquire 24 new clients: an average client acquisition cost of $400.

$9,600 / 24 = $400

Optimisation

It’s obviously important to know how many sales your organisation is making. It’s also nice to know how much each sale is costing you in promotional expenditure.

But, in isolation, this information is not particularly useful.

What you need to be able to do, is compare your actual performance with your optimal performance.

A common mistake in process management is to establish absolutes as targets. For example, it would be tempting to assume that the objective of your sales process is to generate as many sales as possible, for the lowest possible promotional expenditure.

The reality is that sales and promotional expenditure are interrelated. (You can’t have one without the other.) A singular focus on either maximising sales or minimising promotional costs is likely to sub-optimise the performance of your sales process. The key is to determine the optimal relationship between promotional expenditure and sales.

Accordingly, you now need to determine the optimal figures for each of your global KPIs.

Client acquisition rate is easy. Obviously, your optimal figure is determined by the capacity of your production and distribution processes. (There’s little point generating sales that can’t possibly be fulfilled.)

However, your client acquisition cost requires a little more thought.

As the graph below illustrates, as your promotional expenditure increases, the number of clients you acquire should also increase. However, with the increased promotional expenditure, the profitability of each client relationship suffers.

In theory, your optimal client acquisition cost is the point where these two lines intersect.

In practice, it will take some experimentation (and careful measurement) to calculate your optimal acquisition cost.

The starting point for this calculation is the determination of the lifetime value of a client. (Your optimal client acquisition cost will be a percentage of this figure.)

It is difficult to overemphasise the importance of performing this calculation. Without an understanding of the dollar value of a client, it is simply impossible to effectively manage your sales process.

In our experience, because most organisations have no way to value a client relationship, most grossly underestimate the amount that they are prepared to invest in client acquisition.

This under-investment in client acquisition seriously retards the growth of many organisations.

The publishing industry is one industry that does understand the concept of lifetime value. Typically publishers of magazines and other periodicals are prepared to invest at least 100% of the first year’s subscription revenue in order to acquire a new subscriber!

Valuing a client relationship

In financial terms, a client relationship is simply an annuity income stream.

It follows that you can value a client relationship, just as you can value any other kind of annuity (income-producing investment).

You value an annuity using a net present value calculation. (Net present value is the sum of a series of future payments, discounted for the cost of capital.)

To calculate the lifetime value of a client, determine the gross profit you earn in an average year from an average client, then multiply this by the figure in the annuity table below that corresponds to the number of years you retain this average client.

This table shows how much a series of $1 payments, to be paid at the end of each year for a specified number of years into the future, is currently worth, with interest at different rates, compounded annually. In other words, the table shows what you should be willing to pay, today, in order to receive a certain series of future payments of $1 each.

As you already know, Correlex sells a consulting service for $450 a month.

To calculate the lifetime value of an average client, it must multiply the gross profit in a transaction by the number of times Correlex transacts with an average client over the lifetime of the client relationship.

On average, Correlex retains a client for three years. Its gross profit is 70% (or $315).

It chooses to account for cost of capital at its overdraft rate, 9%.

Correlex calculates the net present value of a client relationship by first calculating its annual gross profit, and then multiplying this figure by the appropriate multiplier from the table above:

Average annual gross profit: $450 x 12 x 70% = $3,780

Net present value: $3,780 x $2.53 = $9,563

The CEO of Correlex is surprised to see just how valuable a client relationship is.

Prior to performing this calculation, he was considering reducing his promotional expenditure ($400 per client seemed like a lot — especially for an engineering firm).

Now, however, he suspects that he has been underspending on client acquisition!

Accordingly, he decides to set his optimal client acquisition cost at a (conservative) $900. He also resolves to watch his KPIs carefully and review this figure in six months’ time.

As explained, your optimal client acquisition cost will be a percentage of the lifetime value of a client. The actual percentage will depend on the fixed costs associated with your sales process (and your sales volume). It will almost certainly be more than 10%. It may even be as high as 50%.

(While the idea of investing 50% of the lifetime value of a client in client acquisition may seem ludicrous, it’s important to remember that, once an organisation has passed its break-even point, it’s effectively enjoying a 100% return on this promotional expenditure. Try earning that in the bank!)

Before he can finalise his global performance indicators, the CEO of Correlex must determine his optimal client acquisition rate. A quick call to his operations manager confirms that Correlex is capable handling four new clients a month.

These indicators provide Correlex’s CEO with an overview of the relationship between his sales process’s inputs and outputs.

His suspicion that he is underspending on promotion is supported by the fact that Correlex is growing at 50% of its optimal rate.

Because your global KPIs are longer-term indicators, they do not enable you to manage your sales process on a day-to-day basis.

To do this, we need to divide your sales process into its key components, and then devise a set of performance indicators for each.

You’ll remember that a sales process consists of three components:

• Relationship acquisition
• Relationship management
• Opportunity management

Let’s start with the final component of the sales process and work backwards.

Opportunity management

The objective of your opportunity management process is to convert sales opportunities into sales (remember, we’re assuming that a sale is a new client). This process will generally involve salespeople, operating either in the field, or from a call centre (or both).

A sales opportunity is typically an expression of interest in a specific product or service, generated as a result of your opportunity management process. (Sales opportunities are often called leads.)

As with your sales process as a whole, we are primarily interested in the throughput and the productivity of your opportunity management process.

Accordingly, we will begin by measuring:

• Client acquisition rate (clients per month)
• Opportunity conversion rate (sales/opportunities x 100)

Of course, the throughput of the opportunity management process (client acquisition rate) will be identical to the throughput of your sales process as a whole.

Your conversion rate is the percentage of sales opportunities that convert into sales.

The CEO of Correlex already knows his client acquisition rate.

What he doesn’t know, is how many sales opportunities his consultants require in order to make each sale.

A survey of his consultants’ sales figures indicates that, on average, his consultants convert one in every 2.9 sales opportunities into sales. (Accordingly, his conversion rate is 35%.)

But these sales figures also reveal an interesting phenomenon. Correlex’s CEO observes that conversion rates vary considerably from consultant to consultant. He also notices that there seems to be an inverse relationship between acquisition rate and conversion rate for individual consultants.

In other words, the consultants who acquire the most new clients tend not to have the highest conversion rates, and visa versa.

He wonders why …

Optimising conversion rates

Contrary to popular belief, the primary influencer of conversion rate is not the skill of salespeople.

Rather, it is the design of the opportunity management process.

In our experience, opportunity management processes are best designed with a view to minimising the time between the emergence of a sales opportunity and closure of that opportunity (a sales opportunity is closed when it is won, lost or abandoned).

Increasing the throughput of a sales process may result in lower conversion rates, but this is not necessarily a bad thing!

Sales managers typically manage their salespeople as if a conversion rate of 100% is achievable.

In reality, 100% is rarely an optimal conversion rate. The reason is that, as conversion rates go up, throughput goes down.

Ask yourself, which would you prefer: a salesperson who conducts 5 appointments a day, with a 40% conversion rate; or a salesperson who conducts 3 (highly qualified) appointments a week, with a 95% conversion rate? (Hint: salesperson A generates 10 sales a week, where salesperson B generates less than 3.)

You can take the following steps to increase the throughput of your sales process:

• Break the opportunity management process into a number of logical steps.
• Ensure all sales opportunities follow the same process.
• At each step in the opportunity management process, be sure to up-sell to the next step.
• Actively manage open opportunities.

You can manage individual (open) opportunities with a simple tabular report, like the one below. Normally, a sales team will work through this report in its weekly sales meeting. The key indicators to watch are the number of open opportunities and average days open. (If you sell a number of products with different price points, you may prefer to monitor the dollar value of opportunities).

Most CRM systems produce such a report; alternatively, you can create your own in Excel.

You can also use the weighted value and target close data from this report to produce sales forecasts.

Correlex’s CEO reviews his consultants’ differing opportunity management processes. He identifies the consultant with the most efficient process and resolves to benchmark this process and make it the organisational standard.

This new benchmark calls for a conversion rate of 25% and an average days open of 45 days.

From these figures, he calculates that, at any one point in time, his organisation should have one and a half months’ worth of open opportunities:

Optimal monthly sales: 4

Opportunities required to make 10 sales: 4 / 25% = 16

Average days open: 45 (5 months)

Optimal open opportunities: 16 x 5 = 24

Relationship management

The objective of your relationship management process is to generate a steady stream of sales opportunities from your relationships under management.

We do this by creating an automated communication program. This is a sequence of ongoing communications, where each communication is designed to add value to these relationships.

A typical communication program will consist of a monthly e-mail newsletter and bimonthly seminars or workshops.

We tend to take an indirect (longer-term) approach to the generation of sales opportunities. Our experience is that, if you can design the relationship management process to position your organisation as the leader in its particular field, sales opportunities will be forthcoming.

Events and other activities can be used to stimulate the flow of activities, but on many occasions, these activities will simply time-shift the emergence of opportunities — rather than creating opportunities you wouldn’t otherwise have received.

You need to balance your need for sales opportunities against the requirement to add value to the relationships under your custodianship. There is a danger that, if you design your communications specifically to maximise the flow of sales opportunities, you may compromise the integrity of these relationships.

As with our other processes, we are primarily interested in monitoring throughput and productivity. Accordingly, our KPIs are as follows:

• Opportunities per month.
• Opportunity realisation rate (monthly opportunities/relationships).

Opportunity realisation rate advises you of the correlation between the number of relationships you have under management (the size of your database) and the number of sales opportunities these relationships produce each month.

Correlex has 1,500 contacts on its database. Because all of these contacts are recipients of Correlex’s monthly e-mail newsletter, it referrs to them as subscribers.

On average, Correlex receives 12 sales opportunities a month from its subscriber database.

These 12 opportunities represent an opportunity realisation rate of 0.8%:

12 / 1,500 x 100 = 0.8%

In order to increase the flow of sales opportunities to the 16 per month required, Correlex’s CEO realises he must acquire an additional 500 subscribers:

16 / 0.8% = 2,000

Relationship acquisition

The objective of your relationship acquisition process is to acquire a steady stream of relationships with potential clients and centres of influence.

The input into this process is the investment in your relationship-acquisition campaigns (remember, we’re only tracking variable costs). The output is obviously the relationships (or subscribers) you add to your database.

The precise management of this process is critical for two reasons:

• The flow of inbound opportunities is the key determinate of the throughput of the process as a whole.
• In the absence of an objective performance measure, there is a high probability that you will either under- or over-spend on promotion.

Our performance indicators for this process are:

• Relationship acquisition rate (relationships per month).
• Relationship acquisition cost (cost per relationship).

Relationship acquisition cost is calculated by dividing the variable cost of promotional campaigns by the number of new relationships acquired by those campaigns.

As with our global KPIs, these indicators don’t mean much until we can compare actual and optimal figures.

Optimising your relationship acquisition process

The calculation of your optimal relationship acquisition rate is easy. This figure is determined by:

• Your target database size.
• Your availability of promotional funds.
• The capacity of your relationship acquisition process.

The calculation of your optimal relationship acquisition cost requires a little more thought.

The amount that you are prepared to spend in order to acquire a new relationship must obviously relate to the value of such a relationship.

But how can you value one more name on your database?

The solution is to value relationships using exactly the same methodology we used to value clients.

Your database of subscribers provides you with a flow of sales opportunities.

You can value a sales opportunity by discounting your optimal client acquisition cost for your conversion rate. (In other words, if your conversion rate is 10%, a sales opportunity is worth 10% of your optimal client acquisition cost.)

Accordingly, to value one new subscriber, all you have to do is calculate the odds of that subscriber becoming a client over the life of their relationship with you.

While you can easily calculate the life of a client relationship, it’s a little harder to calculate the life of a subscriber. In our experience, it’s rare for subscribers to unsubscribe from our automated communications program.

For this reason, we arbitrarily choose to value subscribers over the same lifespan as clients. Accordingly, if the life of an average client is three years, we value subscribers over this same period.

The CEO of Correlex is prepared to invest $900 to acquire a new client.

Because his optimal conversion rate is 25%, a sales opportunity is worth $225.

$900 x 25% = $225

He knows that his automated communication program provides him with an opportunity realisation rate of 0.8% per month. Or, to put it another way, for each subscriber on his database, he will receive 0.8% of a new sales opportunity each month.

Because he arbitrarily decides to value subscribers over a three-year period (36 months), Correlex’s CEO can calculate that there is a 28.8% likelihood of a new subscriber turning into a client over this period.

0.8% x 36 = 28.8%

If 28.8% of subscribers become clients, it follows that a new subscriber is worth $64.80:

$225 x 28.8% = $64.80

Therefore, this $64.80 is Correlex’s optimal relationship acquisition cost.

Correlex’s CEO decides to set his optimal relationships acquisition rate at 85 per month. This will allow him to easily hit his target of 2,000 subscribers within 12 months. (Even accounting for a particularly conservative unsubscribe rate of 42 a month.)