The Blog

  • April 4, 2012
  • Why CRM Works

    I have great admiration for the work of Daniel Kahneman.  He’s an Israeli psychologist whose work with the late Amos Tversky won the Nobel Prize in Economics in 2002 even though neither was obviously an economist.  The pair built the foundation for behavioral economics and did early work on framing, or how we perceive issues based on what other information surrounds them.

    Kahneman’s work — the part I am most fascinated with — focuses on how we think and his new book Thinking Fast and Slow summarizes his work and that of others in the field and it has significant implications for our daily lives as well as many applications within CRM.

    Kahneman postulates two systems, or ways of thinking, which he calls System 1 and System 2.  The systems are really metaphors for how the brain works — they don’t show up on an MRI.  System 1 supports thinking quickly, reflexively and in the moment.  It is the snap judgment or gut instinct that evolved to keep us safe from hungry lions and we rely on it still.  System 2 is what many people would call “real” thinking.  It supports the number crunching we do to get quantitative answers, and other deliberate work we do whether or not it involves algebra.

    Both systems likely evolved on the savannah.  When confronted by a predator System 1 was most responsible for fleeing the scene.  System 2 would have been more useful for an individual trying to find the best place to cross a river — not too deep and no crocodiles, please.  You get the idea.

    Kahneman shows through experimentation how the systems work and how we may prefer to let System 1 do the heavy lifting even when it is out of its depth because System 2 involves real work and the brain would rather not engage.  Here’s an example.  A bat and a ball cost $1.10 as a set.  Individually the bat costs $1 more than the ball, what does each cost?

    According to Kahneman for most people System 1 springs into action concluding that the bat costs $1 and the ball $0.10.  Did you get this answer?  It happens to be wrong because $1 is only ninety cents more than ten cents.  If the ball costs ten cents then the bat would have to cost $1.10 which would make the set $1.20 not $1.10.  If you do a little algebra you discover that the bat costs $1.05 and the ball can be bought for a nickel thus conforming to both conditions.

    For many situations the difference between an answer of ninety cents and $1.05 is a difference without a distinction.  But a willingness to accept the wrong answer tells a lot about our brains and CRM use too and the propensity some of us still have to blow off CRM as too much work.  How many times have you heard this regarding SFA, “It doesn’t do me any good.  It’s just extra work because I’m just collecting data for my boss.”?

    I hear it often.  In fact some emerging companies use this as a reason that their products are superior to what’s on the market right now.  Their products are easier to use, less invasive they say.  But forecasting results tell a very different story.  These new vendors claim to be better because they, in effect, don’t ask the user to think and in the process they perpetuate wrong answers.  Thus the road to CRM enlightenment is pockmarked with potholes of ignorance.

    Part of each one of us prefers the ignorance of the wrong answer because System 1 took care of it and we didn’t have to engage System 2.  The ninety-cent answer gets us close enough — that is until we go to the store and discover that we don’t have enough money.

    For a long time there’s been anecdotal information about CRM’s benefits, which has been unconvincing to at least some people.  A diminishing number of System 1 types assume it’s too much work while others might see it as the path of least resistance — two valid System 1 responses.  At the same time though, some System 2 types think there isn’t enough proof and want clear ROI numbers, which are hard to get.  Actually the System 2 types are really just engaging System 1 by expecting someone else to do the work.  It can get complicated.

    The advent of analytics as a part of the CRM suite was supposed to shed new light on CRM’s efficacy and some clear information is now coming to the fore.  Last week I spoke with Swayne Hill a founder of Cloud9 Analytics and the SVP, Global Field Operations.  He’s started blogging about his experience using analytics in selling and the findings are very interesting.

    From his writings I can say that one of the knocks against CRM is valid, it is used to capture data for future analysis, but for a long time that analysis was performed by many managers operating in System 1 mode.  They took the forecast and added their gut feel for the data often promoting some deals and de-emphasizing others.  Despite this effort though, our data suggests that most sales forecasts are profoundly unreliable.

    To take a System 2 approach with a sales forecast almost demands analytics because there is simply too much data for a human brain to crunch in a timely way.  So using his own analytics including a new forecasting tool, Hill began analyzing his company’s performance.  He freely admits in one post that the company missed its forecast a couple of months running which got him and his peers very interested in figuring out why.  Indeed, a company that specializes in sales forecasting has to both eat its own dog food and benefit from it.

    In short order Hill discovered that his sales reps were three times more likely to win a deal when the VP of sales for the prospect was involved in the decision.  That sounds like a no-brainer but it isn’t.  Lots of large organizations have sales operations groups and often they can make the purchase decision for a product like this, especially if it can be shown to save money (that’s the emphasis in operations).  That’s basically a System 1 decision.

    But it’s the sales VPs who can drive a decision if they think it really will boost sales.  So while the operations side of the house might not buy if the ROI isn’t great enough, the VP of sales might be more willing to make the purchase if a vendor can show a modest performance improvement over time and that is a System 2 decision.

    Hill’s discovery led to a metric, a sales milestone that is now built into the Cloud9 internal sales process.  If for some reason a deal gets to forecast stage without the customer’s sales VP being involved the forecast it is marked down for that reason.  So ironically, and very interestingly, this new milestone has become a System 1 decision point.

    Hill’s blog is full of observations like this and I think he’s right to focus on ensuring data collection in sales.  I think every sales organization can go through an exercise like Cloud9 and come out the other end with valuable new insights into how and why it is successful.  But that analysis is a System 2 activity that will ultimately drive System 1 activity.  We need to challenge ourselves to not take the easy route of what’s been done before, to fully engage our minds.

    Finally — and I know I’ve said this before — as important as data is, it is useless without analytics to turn it into real information.  Hill had plenty of data about the importance of the VP of sales but sitting in the database it was diffuse and useless.  When it was organized by analytics it became information, which was, in turn, converted into knowledge that future System 1 types can leverage.

    CRM works and it is worth doing because it gives us the tools to engage in System 2 work to make ourselves more productive.  If we’re simply relying on data capture and System 1 cogitation, our results will be mediocre.  But there’s nothing wrong with using a System 1 shortcut once you know it’s based on knowledge.

    Systems 1 and 2 play into social CRM strategies as well and I will be writing more on the subject soon.

    Published: 12 years ago


    • April 6th, 2012 at 8:45 am    

      Terrific article. Your comment about SFA gave me flashbacks. Very true. And, not my point, but to that comment, I always wondered. “Why wouldn’t you want the same information as your boss? It’s your business.”

      Anyway, I probably have to go back and read a couple more times. But, to help me out, can you walk me through a point I’m stuck on.

      In your Cloud9 example, I’m confused as to how the Sales VP involvement metric is a System 1 decision point. If it is based on quantifiable evidence, derived through an analytical method, wouldn’t that be a system 2 decision? Or, is it the fact that, once the analytics proves the assumption to be true, it no longer requires analysis at each decision point, so therefore moves from system 2 to system 1 because you’re now accepting it as fact without additional analysis?

      I hope my question made sense. And I hope you can sort this out for me.


      • April 6th, 2012 at 8:54 am    

        Hi Barry,

        You are spot on. Once SOMEONE has done the thinking for a metric, everyone else can just use it as received wisdom. That’s the power and beauty of a metric, which I think at this point becomes a key performance indicator KPI. KPIs are like business enzymes that speed up a process or enable it to be performed with less energy input. Thanks for your question.


    Speak Up

    You must be logged in to post a comment.