The recent Gartner report on ‘Combine Predictive and Prescriptive Analytics for Better Decision Making’, we believe makes it clear that the future will not necessarily be won by those who know what will likely happen. It’ll be won by those who know what to do about it.
Clarity In Vision
We feel the report’s strength is how it goes back to basics. While the concept of bringing together predictive and prescriptive analytics is nothing new, these terms have blurred over time and become almost interchangeable. This report tells us why it’s important to pay attention to the differences as we face our current challenges.
Our complex and uncertain world cannot be neatly summarized by a single variable, technique, or methodology. Certainly, we can no longer rely on outdated Business Intelligence tools applying only descriptive and diagnostic analytics. We need a purpose-built problem-solving AI-driven toolkit that applies both predictive and prescriptive analytics so people can have a quick overview of a situation before acting.
“We need a purpose-built problem-solving AI-driven toolkit that applies both predictive and prescriptive analytics so people can have a quick overview of a situation before acting.”
Clarity In Clear Actions
Yes, it’s all about taking action. If you think about it, predicting a problem is a very different value proposition than something telling you what to do about a problem. This is especially true for a world with a chronic shortage of people with the skills to do the actual work society depends on.
You’re not going to make friends by pointing out an issue and then saying, “Good luck with that.” The problem will likely then just be ignored because there are more immediate fires that need putting out.
The action people should be taking is to solve the problem, not just figure out what the problem is. In this way, prescriptive solutions allow experts to be experts and solve the problem they are experts in solving – without having to become data experts.
With prescriptive, you get rid of all the noise so your workforce can focus on what’s important: fixing the stuff that needs fixing ASAP.
Price Versus Value
But we also all know that cash is king and prescriptive solutions almost always have a higher price tag than lower-value solutions. But when you look deeper, the math changes everything when you factor in the costs associated with production loss and the gains from avoiding downtime, waste, and overtime – along with how you can manage your spare parts and logistics. Prescriptive solutions pay for themselves over and over again.
Certainly, once you’ve seen prescriptive in action, there’s no turning back. The value is obvious. Yet we still hear the conversation: “Hey, let’s downgrade here to this cheaper predictive solution.”
Why fight progress?
Predicting The End Of Prediction
The report presents one case study that we believe bypasses predictive altogether:
By deploying prescriptive analytics in the form of an optimization model, a global snack-food manufacturer was able to consider multiple variables – “including various chip size, bag size, fryer speeds, temperatures, seasoning loop capacity, the cost of input resources, and cost of shipping” – to find the best way to allocate their produced chips to multiple distribution centers.
As a result, the company enjoyed a $300,000 per week reduction in cost – these are the sort of numbers that make sense to anyone.
All About Trust
However, whatever prescriptive solutions you go for, you need to trust the people who provide them.
As with doctors, you’re only going to take their advice – and prescription – if you trust their expertise. Conversely, I am not going to trust the recommendations of a shopping platform, if I don’t trust that the company is truly out for my better interests.
And the same is true for the SaaS world: prescriptive only works when people trust you. By providing predictive insights to drive prescriptive actions, it’s possible to eliminate unnecessary downtime, increase capacity and yield, optimize asset care, and reduce waste, loss, and emissions – with a 99 percent success rate.