A couple of days ago, I hosted a simple poll on LinkedIn. I feel the need to stress that the poll was not the result of extensive research, nor did I consider if the questions might be leading. Here are the questions and response percentages:
What’s next for commission software?
- Integrated forecasting with AI – 44%
- T&Q visualization – 7%
- Real-time calculations at scale – 26%
- I wish it just worked – 22%
While there weren’t thousands of people that responded, I thought the number, who responded, and where they responded was quite telling.
The fact that I have been around for some time now in the commission’s industry and have only really seen two real, substantive changes, is telling also. Those two changes were the move from on-premise to cloud computing which caused a re-architecture of most software’s that were existing prior to the cloud boom, and the other was dashboard visualization.
So here we are, 22 years from when I started in this industry and to me it seems little has changed. There is a thirst for change, by evidence of recent investments in the industry whether it be capital into newcomers or new rounds and purchases by incumbents. So there is a desire for change and growth. I’m going to put the three ideas from the questions I posed into some analysis on where companies could take these ideas and why we don’t have anything ready for market yet. The last one, I wish it just worked, well, that I can detail too.
First, integrated sales forecasting with AI. This is a topic I have been giving a lot of thought to thought. For those that are familiar with the basics of AI, questions such as what type of algorithm, what would you use as variables, how to cleanse the data, and how it all wraps up together, to give you something better than a linear analysis have been going round my head.
I also gave some thought as to why so many people voted for this, what business driver is behind it, why would people like to see this functionality. My (educated) guess is that forecasting is utterly inaccurate and unreliable. Taking the benefits of AI and weaving into that forecasting function has to improve it. It needs improvement. If I roll forward to the business benefit of accurate forecasting I can see why the votes were cast.
Looking at the technology here for a while, let’s assume the algorithm is something like MCMC and we will have both internal and external variables. The variables we will use are prior year sales by product or service line, existing capacity and capability with future production expectation, S&M allocated to spend for last year, external demand projections by industry vs current, weather patterns in Northern Ireland.
I think you get the idea, but what I am looking for is a system that lets me pick the algorithm, determine the data and clean it prior to entry. This should have a simple stepped wizard with the ability to refine the projections on a monthly/period x basis. I know what such functionality needs to include. I’m not sure who has this on their roadmap.
T&Q visualization is quite simple. I want something that is better than letting me spread numbers up, down and sideways. I want to be able to drag and drop the expected quota or segment it on zip code/postal code. I want to see how those movements affect a group specifically when I start to take out accounts that might be key accounts. Then I want to see how it can be weighted based on location expectations and if I need to add more headcount or move headcount.
So a simple concept but one that has not yet hit any of the products.
Moving on to real-time calculation at scale is pretty self-explanatory. We are getting close to this, but at this time most of the applications do a batch process calc. Why is it that excel can calc in real-time, but most of these applications can?
Maybe there should be a low/medium/high button so that you can say, hey I’m low today, show me the calc immediately. Or, I’m about to run the mother of all calculations, you better call the power plant? This is one that I have heard over and over from customers as far back as 2005. Specifically, if someone is coming from excel, where this happens, why is it so elusive to this industry. There are some breakthroughs here with some vendors are starting to show this capability.
On to the last point. Just make it work.
Let’s start with why doesn’t it just work?
Well, a lot of time’s that comes back to you.
Is your plan too complex?
Did you go through a valid data cleansing and expectation process?
Did you mock up your reports or just wing it?
I guess this is a subject in its own right, one that my company spends a lot of time with customers resolving (hopefully none of our customers put this down as an answer). The interesting point here is that even with the prospect of new technology none of that matters if the basics are not there.
Any conclusions from this?
I think I know what is possible and my mini-survey showed me what customers want. The action now is for the vendors – show us your roadmaps – many other product genres are adopting these technologies, in order to catch up with SPM. Whilst we wait if anybody is really interested I know a very forward-thinking SPM focused digital consulting company that can help you achieve these objectives.