Sales Management

The Trouble with Sales Science

Forecast 2017

Forecast 2017, the 3rd annual conference defining the future of “sales science” is now history.  I attended this conference, hosted by Base CRM, all day today, September 18, 2017.  I attended it last year as well.  I think it is fantastic and I will attend again next year.  They’re on the right track to something, but I’m not so sure that sales science is that track.

If you’re a sales operations or management executive, and you want to understand the current trajectory of sales technology, then this is a must-attend conference.  If you are wondering about the intersection of data mining, statistics and sales, yes, please attend next year.  That said, this is not, to my mind, a science conference.   Presentation of falsifiable scientific theories of closing sales? No. Peer reviewed papers? No.  Crazy new conjectures on what makes sales “tick”?  No.  New sales applications for cutting edge science, mathematics or computational complexity invented or discovered for other purposes?  No.

Forecast, the annual sales conference, is closer to an 8-hour infomercial for Base CRM, but I don’t see that as a negative.  This is genius marketing, and I applaud Base CRM.  They put on a great event and it’s worth the time and money to attend.  But after decades in sales (and having designed and implemented CRM systems going back to Act 1.0), I can say that Base CRM is missing an incredible opportunity to found a new branch of science.


David Hilbert inventor of Hilbert space

I can only speak personally about my scientific effort at Salesphase, which is focused on high-value, long sales cycle complex B2B deals.  But I have to believe there exists a potential to bring together the best geek minds on the science of sales.  Maybe it could be an off-shoot or sub-conference of Forecast.  I hope I’m not alone in this thinking, but take Salesphase as an example: It posits that only a fraction of the information important to closing complex deals is ever captured by a CRM program or sales automation tool.  Not only that, but the data it fails to capture is far more important to closing deals than what it does capture.  But to fully flush out the Salespahse conjecture, and its value to sales software applications, requires an understanding of complex Hilbert space, vector calculus and complexity theory.

Base CRM is a company uniquely situated to take advantage of incredible computational techniques that have never before been applied in the realm of sales.  Techniques that (metaphorically speaking) replace the current beads on an abacus and notches on a stick with quantum field theory.  I’m not exaggerating either:  The mathematics behind CRM and sales automation tools is how old?  At best 330 years old, as in Newton’s Principia Mathematica published in 1687.


Newton, co-inventor of the Calculus

Babylonian math

Babylonian √2

At worst we’re talking 3,617+ Babylonian lunisolar years old.  The clay tablet pictured at right is a Babylonian calculation of the square root of two. At least they were thinking about irrational numbers, which is on the scale of the floating point mathematics used in CRM data analytics.   But what about imaginary numbers?  They can represent valid non-deterministic states of a real system–states which cannot be directly measured.  This has interesting applications to the possible states of customer systems.


If a sales application took advantage of the power of Hilbert Space to model complex sales, that would move us up to the first decade of the 20th century.  Throw in Topology and Braid Theory and we could even push sales science into the Cold War era!  Salesphase is topological in some aspects, that is, it can be represented as a geometrization of a new sales calculus.  At least it uses mathematical concepts from the 1950’s.  I’m confident intelligent young people could get sales into the 21st century of mathematical concepts.

Here’s the key problem: most of the data collected in CRM systems has little to do with successfully closing a big or complex B2B deal.  Does logging a voice message left for a buyer count as “data”? Yes.  Does it count as data relevant to closing a deal?  No.  Does a data point representing a sales presentation at a customer’s office accurately tell us about the health, complexity or probability of closing a big sale?  Maybe.  But, I’m certain there is a wide standard of deviation on the usefulness of this data.  For better or worse, sales reps, in my experience, tend to be either over optimistic or too pessimistic.  Either overstating a chance of closing, or sandbagging a sure thing as a “maybe.”

CRM “simple math” drove me crazy as a senior Sales Executive at 3M Company.  At that time, for a hobby, I was writing science book reviews for a literary website, and I thought, “Why aren’t we using these amazing current mathematical/scientific advances to improve sales performance?”  I personally used just a little of what I’d learned to crack the enigma that was Boeing Commercial Airplane’s Wichita Division (now Spirit Aerosystems).  Ten years later though, I’m still asking the same question: why are sales organizations using the most advanced mathematical and scientific techniques?  My answer so far is, “the language barrier.” The vocabulary that scientists and mathematicians use has no meaning to sales software developers.  I would enjoy being a translator in this case, I think a lot could be accomplished.

Unfortunately, the “sales science” behind the latest software tools is more limited to statistical data analysis of information gathered by CRM programs.  Despite what I am saying, these efforts WILL be successful.  That’s because doing a better job analyzing mounds of information is guaranteed to improve sales performance.  But is it the optimum method in all cases?  No.  Anyone with a computer science degree knows that for big problems, search-sort is less efficient than similar sized decision problems.  Salesphase is more like a set of decision problems, it does not require mounds of often stale contact data and sales history to predict the next best move, or the probability of closure, or the “state” of a customer with respect to signing a contract for your product.

I have always maintained that much of CRM data is either to “explain what you’ve been up to” to a manager, or to funnel information to marketers.  Of course, that’s not always the case.  I once solved a major customer nightmare (involving the manufacture of specialty optical fiber) using Siebel CRM 7.5.2. It worked because there were so many people involved over such a long period of time that I could not rely on my memory or email.  I probably could have used an Excel spreadsheet, but I was part of the Siebel implementation team for my small corporate business unit, and I wanted to demonstrate to my boss what CRM could accomplish.  3M Optical Components won a supplier award from Honeywell Defense & Space as a result.  So, don’t get me wrong.  I’m pro-CRM.

During Forecast 2017, the closest thing that came to science was the presentation by CEO/Founder, Amit Bendov.  Why was it the best? Because it wasn’t anecdotal.  It wasn’t the perspectives of a VP of Sales, a Chief Revenue Officer, or an executive panel talking about inspiring sales teams.  Amit presented the outcomes of data generated by a what appeared to be proper study of 3 million minutes of recorded sales meetings/calls.  It revealed important correlations with “success,”  where success is defined as “moving a deal forward.”  My ears perked up for this.  The average length of a successful sales presentation is 46 minutes; the longest sales rep “monologue” in a successful sales meeting is 76 seconds; in the second half of a demo, the amount of times dialogue switches between sales rep and customer increases 36%; finally price isn’t discussed in detail until 38-46 minutes into an average successful presentation.  Clearly, the number 6 plays some mystical role in sales success!  I loved this.  It’s a bunch of useful benchmarks.  How might successful sales science theories be measured?  By comparing results of sales outcomes using a new theory with the benchmarks that represent a spectrum of traditional sales techniques (as represented by Gong’s sales meeting data set).

I see great things for the growth of the Forecast annual conference.  In a sales technology industry dominated by, Inc., this conference is a welcome respite.  But, please, if there’s anyone else interested in building a community for rigorous sales science, contact me.  –John Clark, john @ salesphase . com. 



Is Gauge Variance Messing With Your Sales Team?

Discrete Scale Invariance

Discrete Scale Invariance

Let’s tap into the abstract concept of gauge variance to distinguish a particular type of common sales team dysfunction. First, bare with me as I explain what this gauge variance thing is all about. I will start with the exact opposite property, gauge invariance because the most fundamental aspects of the universe possesses this property–more often called gauge symmetry.  This extremely abstract property is difficult to adequately describe here, but its effects are very important to us. For example, we literally see things because electromagnetism (light) is a necessary consequence of the gauge invariance of the electron field.  Because the dynamics of electrons are constrained by gauge invariance, it necessarily follows (thanks to some mind numbing graduate level mathematics), that photons of light MUST exist to transmit energy (some would say “information”) between electrons. I for one am very happy Mother Nature requires gauge invariance because, among other things, it enables me to see Facebook status updates alerting me to stomach ailments of “friends” I haven’t spoken to in 25 years.

Setting aside Facebook and electromagnetism, think of gauge invariance like this: if I measure the value of a sales deal in dollars and my customer measures the value in euros, we do NOT get different results in terms of value. There is always a publicly available market-driven currency conversion factor which we can use to verify that we both measured the same value for a deal regardless of our choice of currency (our choice of gauge). Now contrast this with measuring something like the coastline of California.

If I measure the beautifully rugged California coastline with a yardstick and my counterpart measures it with a ten-foot pole, will we get the same result?  No!  The concept of “length” in this case fails gauge invariance–it is gauge variant.  It turns out that the length of any randomly uneven surface we measure can grow to infinity as the choice of measuring stick gets shorter.  As our choice of ruler gets smaller, our measurement picks up ever smaller variations in what makes up the “edge of California.”  As we would expect, this is related to resolution.  As you zoom in on the coast, you continually see more detail of the actual coastline.  In essence, measuring a coastline is scale dependent.  Ask me how long the coastline of California is and I will ask you how big is your ruler?

Now, my proposition with respect to sales is that sales teams often measure deals (informally or formally) in ways that are scale dependent. To say the same thing in different words, sales analysis tends to be gauge variant.  It is the differences in choice of gauge which lead to misunderstanding, disagreement and wasted effort.  For example, I found during my 15+ years in sales that sales management and marketing tend to work at the macro scale to set pricing, make national sales predictions, etc.  Sales reps conversely tend to work at the micro scale of individual customer agents, budgets and local competitive conditions.  It doesn’t take a lesson in physics to intuitively understand that this leads to mismatched results and disagreements about what is happening or what is possible.  Nevertheless,  I prefer the rigor of math and science to stimulate my thinking and analysis of age-0ld sales problems.

Gauge variance whether in physics or sales is a constraint that a theory must have to remain consistent (only sensible answers and no contradictions).  In the sales world, this can be intelligently worked around or swept under a carpet (preferably up in the marketing department where no one will look), but how often does that currently happen?  How often did I see, or participate in, ground-up sales estimates that started with local competitive and customer conditions and then rolled up logically to the larger scale of national sales budgets and pricing decisions?   How about never. It might get talked about, but it never actually happened.  Budgets, pricing, quotas, etc. were all top down starting from CEO/CFO expectations for the division, and from division leaders to business unit directors.  Business unit directors would then work feverishly with sales and marketing management to come up with plans that would be acceptable “upstairs”.  Finally, at the annual division sales meeting that kicked off each new year, I’d get a territory expectations report that made me want to shout, “You want WHAT when?”  This is only one example of where the problem of gauge variance starts to rear its ugly head in sales departments.  At each point in the process, from the CEO down to the sales rep, there are dozens of conscious and unconscious choices of gauge that fail gauge invariance and lead to unexpected results, disagreements, confusion and general sales friction.  Tackling this issue might be too grand a vision to start.  What about within sales departments?  Why not acknowledge problems of gauge choice upfront?

Typically a sales manager will inquire how a rep is coming along in terms of hitting sales targets.  If the rep is at or above quota and feels confident, then the response will be “No problem!”  But if there is a problem, the rep typically responds with a stream of detailed customer “issues” that are creating barriers to success.  The manager’s choice of scale is likely “percent of quota” or “percent of budgeted revenue dollars.”  The reps response is based on some completely different measure such as customer specs, customer budgets, etc.  The question is who is right and who is wrong?

It’s hard to tell wrong from right when the choice of gauge is not explicit.  The manager can be right that the sales rep is under-performing budget, while the sales rep is also right that according to his choice of gauge, he is presently achieving super-human progress against insurmountable odds. What would be better is to have ways to measure performance at specified scales and then a way to connect those results across differing scales.  Great sales managers are intuitively good at changes of scale and how to connect between them.  Once a sales manager sees a problem at one scale, she changes her perspective and engages with the rep at the scale where a problem is occurring.  Conversely, a great sales rep who is under performing plan will FIRST acknowledge that according to the choice of sales management gauge, his numbers are “in trouble.”  Then he openly leads his manager into a discussion based on a different scale and asks for guidance and help AT THAT SCALE, which when rolled up to another scale will make a positive difference.

From the beginning of inception I have developed SalesPhase to avoid scale dependent problems that occur in sales measurements.  How?  By taking advantage of something known mathematically as discrete scale invariance, which in this discussion might be better termed, discrete gauge invariance.  By carefully choosing a gauge that works at multiple (but fixed) scales important to corporate management, sales management and sales reps, a significant amount of sales friction can be avoided.  When management and reps share a language of measurement that is discrete gauge invariant, there is far less disagreement and far greater focus on solving the problems at hand.

If  you are interested in discussing the mathematics and science underlying the SalesPhase approach to sales management, tracking and communications, follow me on Twitter or tweet me privately, John Clark @SalesPhase.



Sales Friction & Broken Symmetry

friction“SALES FRICTION” is wasted energy in the form of wasted time and resources during the process of a buyer and seller coming together to close a proposed deal. During 15 years managing complex high value long sales cycle B2B transactions, I found great advantage in finding ways to reduce sales friction. Simply stated, aligning sales efforts with the customer’s buying phases was the key to reducing friction (as happens in any physical system which gains coherence). But, how to do this consistently when starting with imperfect or zero knowledge of the customer’s buying agents and internal processes?

Due to superior knowledge of their sales process, sellers naturally select their internal sales funnel as a “sales gauge.” Unfortunately this is a poor choice of gauge if one is trying to measure the likelihood of a SPECIFIC customer making a SPECIFIC (large or complex) purchase. There are statistical correlations between a vendor’s sales stage and probability of closure, but can we do better than these correlations? Yes, we can.

Salesphase is a methodology that is independent of any particular buying or selling process. The scale and scope of the deal makes no difference. Sounds “universal?” Let’s check:  to claim a “universal solution” to any set of problems (including sales friction), it is well established historically & mathematically that such solution MUST be based on one or more underlying physical or abstract symmetries. “Symmetries” can be difficult to define, and as an aspect of Salesphase is kept securely tucked “under the hood.”  But in the context of the underlying algorithm, symmetries are groups of transformations which leave some property of a system invariant (fixed, unchanged, indistinguishable).

Finding hidden symmetries can be difficult. Customer and vendor organizations and processes often appear specialized, unique and yes, asymmetrical, when compared to one another. Why? Because symmetries can be “broken” by external forces, e.g. culture, terminology, system noise, etc. Broken symmetries, when not purposefully designed into a system, necessarily imply at least two things:

  1. There exists hidden “embedded” useful information; and
  2. Some property or characteristic of the system remains invariant despite observed “local differences.”

Salesphase exploits invariant properties of the selling/buying process to unlock hidden information critical to reducing sales friction and other sales dysfunctions. Salesphase realistically assumes a world of IMPERFECT/INCOMPLETE customer information which can be enhanced as information is learned or modified. Salesphase is a methodology leading naturally to an algorithm, and is not competitive to CRM or data analytics. Instead it is a framework for extending the value of existing tools while reducing total sales effort and increasing the accuracy of sales predictions.

Tweet or follow John Clark, @Salesphase, to discuss.