I just finished reading “The Black Swan”, a book that has been on my list since it came out in 2007 and I highly recommend it, though it is not easy reading. There is a great deal of set up before you get to the whole point of the book in the last 50 pages.
“The Black Swan” is about uncertainty in the real world and the subtitle explains it all: “The Impact of the Highly Improbable” and it is something that I can see affecting CRM and its users on many levels. Highly improbable things happen frequently and they have deep and unpredictable impacts on our world.
Uncertainty is related to risk and randomness, and while we lay people might lump all of them into the same definition, they are different. My understanding of the three is that risk is something to which you can attach a probability like a coin toss landing heads (50/50 or .5). Randomness is less constrained than risk if only because it is too complex to compute — if we flip one hundred coins we might reasonably expect half of them to land heads but we can’t say which ones. Uncertainty is a condition or event that will surprise us, one that is not on the radar. A blue bird, sometimes called an unknown unknown.
The blue bird interests me most and is an important factor for CRM. I have been trying to get my head around some data that I collected earlier this year about sales and forecasting that seems to relate to this. As you might recall, the data showed that the vast majority of sales forecasts (better than 90% of them) are worthless. They are so inaccurate that they can make a coin toss look like the model of precision.
My question: Is this the best we can expect or are there things we might do to improve our forecasts by reducing uncertainty?
The data also show that most sales managers engage in a process of downloading forecasts to spreadsheets in which they massage the data with the intention of improving it. If the dismal forecasting results are the product of an improvement process, it can only suggest that the starting point data is no better (remember GIGO).
All this evidence notwithstanding companies continue to make money more often than not and sales people manage to sell things. And according to Jim Dickey and Barry Trailer nearly sixty percent of them make quota. How can this be? How can we be so profoundly bad at sales forecasting and still manage to sell things?
First, though there is a great deal of uncertainty in sales forecasting, uncertainty is a two-way street. Deals come in that were not forecasted or perhaps not even known about — so-called blue birds — and some deals that appear to be a lock simply evaporate. Anyone who has tried to sell — and forecast — knows this.
Customers do rational things for emotional reasons, the saying goes and the irony to me is that at precisely the moment when we need the customer to act emotionally — to buy a product — we expect that customer to act rationally. If you doubt this then how can you explain attaching a probability to an otherwise emotional decision. It’s not wise to do that so if you are going to forecast using a probability of close, then you have to assume a set of rational expectations. In other words, if we have been through a sales process with the customer — understood the business problem demonstrated how our solution solves the problem and asked for the order — we expect the logical conclusion, a purchase order.
But while the situation might look certain or at least logical to us, we have little or no visibility into the similar process being conducted by our competitors and there lies a great source of uncertainty.
What can be done?
As mentioned above with GIGO, we need to acknowledge that the way we are forecasting is not working; in other words, stop digging the hole we have gotten into. There’s too much uncertainty in a forecast to believe the numbers we generate.
But stopping the excavation will not solve the problem; it will simply prevent the hole from getting deeper. As sales people we can try to eliminate some of the uncertainty in our deals but by definition, we can’t do that. We don’t know what we don’t know. Even if we know everything our competition says and does we have no visibility into whether the stock market will crater the day before we expect the P.O. There is always something.
Nonetheless, a company’s sales usually fluctuate around a level that is near to the level of the goal — few sales teams in aggregate hit the ball out of the park and few get shut out. Uncertainty makes sales a numbers game meaning the more irons you have in the fire the more opportunities you have and the better insulated you are against risk. Notice I said risk and not uncertainty. More opportunities reduce the importance of a single opportunity because there are many ways to make your number, that’s risk. But there is nothing that will insulate your forecast from the remote (we hope) possibility of the stock market cratering or flu breaking out in your customer’s headquarters.
If sales really is a numbers game then it makes sense to have systems that can help you manage big numbers of everything — opportunities, deals you know. More important, it is also essential that we have ways to get as many good opportunities into a pipeline as possible.
For years, we’ve had a discussion about the efficacy of SFA — is it good, is it worth the effort, isn’t it just a management tool? Ironically, there are still companies out there that believe they are too small for SFA or that it doesn’t work. In a few years it will be the same stories with social CRM and by then the companies that failed to adopt SFA will be completely out of business.
Truth is, we need both these days. We need SFA to manage our large data sets and we need social CRM to help us take uncertainty out of deals. If we never use an outbound social CRM tool such as a blog or micro-blog or networking site we would be fine as long as we had some exposure to tools that help us know what customers think in aggregate. For me that’s where the power of social CRM is, it’s in helping us reduce uncertainty by that radical idea of asking customers what they think.