Categories: Sales and Distribution
Tip #6: Long-range forecasting is a math exercise, not a sales task.
We posed this to Kraner: OK, the Ultimate Contract looks like a great set of "reality goggles" to view all the deals that are currently in the hopper. But what about the stuff that's farther out? How do you predict sales for next year or beyond?
"Forecasting that far in the future is something that is done by statisticians," says Kraner, "and sales reps are not qualified to do it. It needs to be something separate. It's a branch of mathematics, it deals with past data, analyzing trends, seasonality -- and as you get into that, there are an infinite number of things you can consider.
"I'd hire someone with business school experience, or I would have someone on my staff to look at things like trends and seasonality.
"It really depends on the length on your sales cycle; you can go out just that far. If your sales cycle is six months, you can ask sales reps to forecast out six months, using the Ultimate Contract or some form of customer verification.
"In a short cycle, like two weeks -- well, in that kind of environment, forecasting is simply not done. For example, Dell's inside call center does not have their reps producing sales forecasts. It's a separate department."
Tip #7: If you need forecasts for funding, use comparables.
There's another time when sales predictions are expected: When a startup is trying to get funding. After all, investors want to see when their big payback is due. How do you get to meaningful numbers if you're in this boat?
"Steve Jobs' answer was, to paraphrase, that all market research is BS," Kraner says. "In the end, most of those numbers are 'swags'-- which, remember, is an acronym for 'scientific wild-ass guesses.'
"It's like the story I told you about WWII weather forecasting: People want those numbers, but the most knowledgeable people realize those numbers are not particularly useful.
"However, if you can find a proxy in the territory that you're looking at, you can estimate how much business you will do. In my business, training, what I discovered is that if I bought the Nightingale list, for everyone who has bought in this zip code or area within the past year, I could estimate what I could sell.
"Let's say I’m looking at patch B, in this zip code: I want to know what it is worth. If I think there is a relationship, I can look at patch A, and see what a successful rep is doing there, then estimate the business that could be done in B.
"But, I have to have a known relationship somewhere. If I were a Cisco reseller looking at what kind of business I could do in Guam, I might look at the ratio of cell phone or data centers or internet providers in some other territory similar to Guam. Then, maybe, I could estimate what I could do in Guam."
Tip #8: Bloated forecasts might hook the VCs, but it won't bring them in the boat.
"If I were hired as a manager by a software company with questionable forecasts," says Kraner, "I would first select the selling system, then deploy and train people, and I'd have a have customer verification system in place.
"I'd then give the entire team 30 days to re-do the forecast. Everything has to come off the pipeline, and it comes back on only if there is a verifiable contract. And this means the forecast number, in dollars, will likely decrease.
"The first time we did this was with a VC-backed company out in the Midwest. The sales VP called me a couple weeks later, in a bit of a panic. The VC guys who'd hired him told him, 'We brought you in to increase sales; did you not understand?'
"Of course, what happened to the forecast was that it had evaporated when the VP did what I said; there was a lot of crap in there. He survived that quarter, and became one of the VCs' favorite guys, because from that point on his numbers were believable.
"How did he survive? Well, he almost didn't -- they almost fired him. He put back a forecast that was about one-fifth of the previous one, but it all closed. In other words, the original subjective forecast was off by about 80 percent."
Tip #9: No matter how good you get, you're unlikely to hit 100 percent.
Even when you use a forecast system based on customer verification, some of your predicted sales won't come through, Kraner says.
"At an organizational level, 80 percent is as close as we've been able to forecast. About 20 percent will not close; no action is taken. But, that doesn't mean the deal is going to your competition. Maybe the company gets bought or acquired. Funding goes away. The person you're dealing with goes away.
"What you have removed from your forecasts is the subjectivity, and what you have remaining is customers who are truly engaged. But it still doesn't turn into a closed sale 100 percent of the time."