What Sales Process Behavior Charts Can Tell You

Dear Readers,

Thank you for the excellent remarks about what process behavior charts can tell you from last week’s blog post. For convenience, I have reproduced the image here (with interpretation below).

Figure 1 - Sales Process Behavior Charts

Figure 1 – Sales Process Behavior Charts

Mark Allen said:

Looks like while they are closing, they are not also prospecting…and unfortunately I see B2Bs do this frequently. This is by far one of the single biggest headaches that drives CEOs crazy as it adds a degree of variability that hits home where it hurts…in cash flow.


Sales Production Bottleneck

In many companies, salespeople are required to be involved in everything, so they are the bottleneck. Managers drive them to “burn both ends of the candle,” holding them accountable for simultaneously keeping their sales funnel full while also closing enough business.

The process chart makes this invisible conflict visible. It shows pretty clear evidence of a bottleneck, which implies that a division of labor might be able to improve the flow of revenue. In other words, if they could set things up so some people were consistently generating leads, others could consistently be closing them.

Dave Hurlbrink made an interesting observation:

I infer from these sales process charts that, although the team made their numbers during the year in question, they wouldn’t make them the following year. Also, I suspect their process is biased toward later-stage selling activities versus early-stage ones.

As it turns out, both of these statements are true, although I’m not sure you could prove it just on the basis of these charts.

Sales Production Flow

The thing is, to understand enough to make reliable projections, you need to understand the flow. To do that, you need a measure of input as well as output.

The big problem in 99% of organizations is they don’t even bother to measure the input; they only have a list of deals on a report they call the “forecast,” with nefarious closing probabilities associated with them.

Bob Atkins pointed out some additional interesting information:

I agree that it looks as though closing is depressing lead rates, which can cause a feast-and-famine cycle. …

Last, the controls limits include 4-6x variation between the lower control and the upper control. This indicates a process that is not really under control at all– and will swing wildly from period to period.

You are right about the feast vs famine cycle, Bob: even if the market is not cyclical, a sales process with a bottleneck like this causes a roller-coaster ride. However, I’m not comfortable with your comment about the process not being in “control,” for several reasons.

Understanding Variation in Sales Production Metrics

First, we are using a process behavior chart intended for analyzing individual variables from an infinite population (an ImR chart), rather than for analyzing subgroups (samples from a large population) as is typically done in manufacturing processes (XbarR chart).

Second, without getting into an explanation of the statistics, suffice it to say that the upper and lower control limits are calculated from the data itself. We are dealing with an extremely small sample size, and all those limits tell us is that the next data points have a 95% probability of staying between the upper and lower “control” limits.

Third, it is important to distinguish a possible equivocation on the word “control” in this context. The behavior of the process is what it is: the “control limit” (a poor name for it, IMHO) is telling us only a characteristic of the process. If data points suddenly appeared above or below them, it would indicate some kind of “special cause” rather than the “normal” variation shown here.

On the other hand, management may need the process to behave in a different manner. For example, they might want the variation reduced. These are two entirely different things. Just because there is a wide gap between the UCL and LCL (upper and lower control limits) does not mean the process is out of control.

Knowing both the mean (average, shown by the green line) AND the degree of variability is crucial to knowing whether you have improved the performance of a metric or not.

Detecting Cyclical Buying and Selling Behavior

David Burton brings up a great point:

Closed orders look to me as though they are driven more by client annual budget (spend it or lose it) pressures, rather than by the sales process/proposition creating urgency.

This is why orders are low at the start of the year and build as the year goes on (then [probably] fall off a cliff again for January).

This market could indeed have been cyclical, driven by customer’s fiscal year and financial requirements. However, it could also be caused by having “trained” customers that they will get a better deal if they wait until year end. (It could be a combination of both.)

So, what does this imply? Does it mean the bottleneck doesn’t really exist?

You won’t know until you test it.

The Need to Smooth out Production Flow

Cherie Durkin points out the value of achieving a smooth flow, if you can:

I see the “close under pressure” in full force. Of course these are inferences but, with sales closing at a higher rate at the end of the month, they are selling with the pressure of the close and perhaps are relying on this technique. That is an issue and one we see in administration in such areas as collections and billings.

The idea is the “push is on!” The downfall is throughput is very poor as we lack the smoothing consistency of the close, which of course smooths the backlog for ops and billing for revenue flow.

This is absolutely true, and the lumpy flow of business (combined with poor forecast accuracy) drives many an operations executive crazy. It ruins cash flow, drives higher costs, and makes it much harder to show consistent profitability.

This kind of sales process cries out for improvement. So, why do companies continue to run this way?

One reason is that they have not seen any other kind of sales process. Often, they do not know how to generate qualified prospects (like everything else, it is left up to the salespeople to worry about). If the marketing department is doing anything at all, the “so-called” leads they provide are often (though not always) considered to be a waste of time.

Another reason is having no means or precedent for “changing” or improving the sales process. No measures, no one who knows how, no one to take the lead, especially since it crosses the marketing and the sales function. Just considering the possibility of doing this can feel like peering into an abyss in some companies.

It doesn’t have to be this way!

Optimizing B2B Internet Sales Processes

Four Case Examples from David Bullock
Thursday June 4, 2009, 3:00pm Eastern

On Thursday of this week, I’ll be interviewing David Bullock, an ex-engineer turned B2B Internet Marketer, and author of “Barack 2.0,” a fascinating and authoritative study of Barack Obama’s use of social media to win the presidency.

David will reveal the fascinating mixture of marketing and selling required for B2B companies to successfully leverage the Internet for lead generation, as well as revenue generation.

Visit Optimizing B2B Internet Sales Processes and reserve your place for this unique event.

I hope you’ll join me on Thursday.

Michael Webb
June 1, 2009

Michael Webb

Michael Webb founded Sales Performance Consultants to create a data-driven alternative to the slogans and shallow impact offered by typical sales training, sales consulting, and CRM companies. Michael helped organize and delivered the keynote speeches for the first conferences ever held on applying Six Sigma to marketing and sales. Connect with me on LinkedIn.

Click Here to Leave a Comment Below
Eero Karjalainen - June 12, 2009 Reply


Your blog is very interesting. Thank You. Have You a list from the most benifical measure for the marketing and sales, where I can use behavior charts?


Eero E.

Michael Webb - June 12, 2009 Reply


The purpose of process measurements is the same whether you apply them to sales and marketing, or to any other process. What you measure, depends on your purpose, on what you are trying to improve.

Generally speaking, getting ANY data about the flow of your leads, opportunities, and deals (rather than just your revenue or order flow) is a major improvement for any business. So, these are ideal candidates for process behavior charts. Beyond that, when you start measuring the quality of your sales opportunities and other variables, the sky is the limit.

For a more detailed explanation of process behavior charts, I recommend anything written by Dr. Donald J. Wheeler. This is a link to his paper “How to Make Process Behavior Charts Work for You:




Rob Howe - July 30, 2010 Reply

I am interested in the impact of autocorrelation with regards to Process Behavior Charts. Monthly or quarterly data is often related to the previous month(s) results. Is that not probematic with IMR or other control charts?

    Michael Webb - August 13, 2010 Reply

    Rob, your question shows you have an advanced knowledge of statistics.

    I asked my friend Bill Bentley, president of Value-Train (www.value-train.com), a six sigma training company, to answer it. Here is what he said:

    “The simplest answer is when you have a process that has a natural cycle, you normalize the cycle out of the data first before control charting it. That’s simple to do and to explain to management.”

    If you need further technical insight around this issue as it relates to IMR or other control charts, I would refer you to Dr. Donald Wheeler, whose website is http://www.spcpress.com. Mr Wheeler is a well known expert in these matters and has written about them extensively.

    Hope that helps!


thinkbetter.ch - April 6, 2014 Reply

I’ll immediately seize your rss as I can not to find your e-mail
subscription hyperlink or newsletter service. Do you have any?

Please allow me know in order that I could subscribe. Thanks.

    Michael Webb - April 25, 2014 Reply

    RSS feeds can be found on the right side of the Contact Us page.


Leave a Comment:

Verified by MonsterInsights