Sales POP - Purveyors of Propserity
Artificial Intelligence, CRM and Sales Prediction
Blog / Sales Management / Nov 7, 2017 / Posted by Nikolaus Kimla / 604 

Artificial Intelligence, CRM and Sales Prediction

0 comments

Continuing our series on artificial intelligence and sales, let’s take a look now specifically at AI as it relates to CRM and sales prediction. As I said before, we could go off in a million different directions in discussing AI, but I’m going to stick with what I know well, which is CRM.

Two very popular buzzwords today are predictive insights and prescriptive insights. But while we’re discussing artificial intelligence, we want to pay particular attention to where AI does and doesn’t make sense in relationship to these two areas.

I’ve already pointed out–and it’s pretty obvious–that AI will continue to become increasingly important in just about every industry. But there is an extreme risk factor when someone attempts to use AI to predict human thinking.

Homo Economicus

Back in the 1920s, economist Joseph Schumpeter left the Austrian School of Economics because he believed–as mainstream economics has now come to believe–that human behavior could be predicted with mathematical equations. The Austrian School, to which I very much subscribe, is highly skeptical of such methodology. In short, you cannot place human beings into an equation and predict how they will feel and act.

The “human being” being so predicted is not a human being at all, but a mathematical construct called Homo Economicus. According to current mainstream economic theory, Homo Economicus will react the same way every single time to various economic stimuli. If Homo Economicus is placed into various algorithms, theory says, then we can use those algorithms to make predictions about human behavior.

The reason this doesn’t work, according to the Austrian School of Economics–and, really, according to logic–is that human action, which is central to any economic system, is not predictable. Every human being takes different actions according to their own desires and thoughts.

Prescribing Sales Behavior

It follows, then, that when it comes to sales, and the way that artificial intelligence is used in sales and CRM, that “prescriptive insight” which would seek to inform a salesperson on how they should behave and what actions they should take, would be risky at best. The claims made for such current AI products are that such AI will leverage deals, allow more cross-selling, reduce sales cycles and leverage campaigns. But given the actuality of human behavior, these are faulty assumptions.

How well do these methods work? How well can a salesperson be put into an algorithm to prescribe behavior? Take a look at the broad statistics: 70 percent of salespeople do not fulfill their quotas. So in the strictest sense, it can only be predicted that they will fail to make quota!

Sales and Complexity

There is a beautiful novel by Hermann Hesse called Das Glasperlenspiel (English: The Glass Bead Game). It takes place far in the future, and in which various factors in all subjects are masterfully related through a game of marbles. It is a highly complex game, given all possible variables. The more variables involved, the more complex it becomes.

Moving this view to sales, today sales has become quite complex. Just within my own industry we have many different kinds of roles. We don’t just have a “salesperson”–we have sales development reps, we have business development reps, we have inside sales reps, and outside sales reps. We have account managers. We have “farmers.” We have customer success managers. All of these are types of salespeople.

Even all of these can be broken down because in each type we have beginners, more experienced reps, the veterans, the underperformers and overperformers.

Right there you have 25 or more variables. And that’s before we take into account different sales territories across international lines (it should be noted that CRM powered by AI is being sold only to enterprises that have such scope, as they are the only ones that can afford to pay for it) which of course introduces a whole other set of variables. A salesperson in Europe is definitely selling differently than a salesperson in the US.

So how could AI and its algorithms possibly take all of this into account? You guessed it: it’s not very possible.

How, then, can artificial intelligence be used in sales and in CRM? We’ll speak to that as we go through this series.

Unpredictable Changes

The most prominent AI CRM product being sold at the moment is sold on the presumption that AI will affect how decisions are being made within organizations, and how to sell. AI will provide greater insight into customers, leveraging big data to identify when they will purchase. AI will forecast purchasing trends and inform promotional activities, and even allow prediction of marketing events.

Right. Aside for the whole human issue discussed earlier, changes in the economy, the markets, and commerce can occur so fast it’s like the wind suddenly changing direction.

Let’s take a look at events that greatly impacted sales, commerce and the economy that nobody predicted. We can start with 9/11, which in just a few hours completely upset the economy of the entire planet for several years to come. It was certainly not predicted. Nor was the recent terror attack in New York City, which also impacts markets and the economy.

In Houston, there were probably plentiful economic activities and events being predicted–right up until Hurricane Harvey, which totally upset all possible predictions. And the hurricane itself, of course, could never have been predicted, right up until the time it was on its way.

There are just too many variables for such predictions to hold enough truth to risk investment in.

What Can Be Done?

In the strictest sense, prediction itself always contains an element of untruth. We cannot really know what will happen until it happens. The only things we can really predict are such events as, “If I throw a rock up into the air, it will fall back to Earth.” But we cannot predict the result of a $1 million investment into marketing, which is one of the major issues we have today with marketing.

As we proceed forward with this series, we’ll discuss what I think should be done with AI, and what is truly possible.

As a final note for this article, there are definitely questions of what could be done and what should be done with AI. As an example that I’m sure everyone can relate to, how do you feel when you call a large company and reach an automated answering system? You’re given such options as, “Press 1 for customer service. Press 2 for sales. Press 3 for…” arghhh!!! You just want to talk to someone! Much of the time you’re so angry that by the time you actually do reach a human, you’re ready to rip their head off. This is an application of AI that most of us–except perhaps the bean-counters in companies that are saving money through its use–wish had never happened.

Where do we go now? Stay with me and let’s discover it together.

Pipeliner CRM allows you to predict customers as closely as possible. Download a free trial now.

About Author

A 30-year veteran of the computer industry, Nikolaus has founded and run several software companies. He and his company uptime iTechnology are the developers of World-Check, a risk intelligence platform eventually sold to Thomson Reuters for $520 million. He is currently the founder and CEO of Pipeliner Sales, Inc., developer and publisher of Pipeliner CRM, the first CRM application aimed squarely at actually empowering salespeople.Also a prolific writer, Nikolaus has authored over 100 ebooks, articles and white papers addressing the subjects of sales management, leadership and sales itself.

This website uses cookies. By continuing to use this website you are giving consent to cooking being used. For information on cookies and how you can disable them, visit our privacy and cookie policy.