(True Story circa 1991)
I have spent the past 30 years building Business Intelligence systems. The very basic idea is that there are some things humans do well, and there are things computers do well. The magic is found when you make them work together.
The management at Mattel had a problem, which is that on any given Christmas, nobody knows what the hot toy is going to be. For all of their years of successes, it’s very hard to predict which product is going to win and which will lose. But some time around November, things get crazy, and the execs had better decide which set of all of their thousands of products should get the green light for full manufacturing blast.
As with every business, especially in consumer markets, making supply meet demand is sometimes art, sometimes science and sometimes pure guesswork. So how can you get 50 managers, each pulling for their own favorites to come to a consensus about what’s best for the company? Well, it’s almost easy. You tell them to put their money where their mouth is.
So I was hired to integrate the list of products into a system that allowed managers to create syndicates and present those syndicates to the execs who would then make wagers (with votes) to gather a consensus. So you had maybe 300 different Barbies 400 different Hot Wheels and yoyos and slime and monsters and who knows what else. I got all of the product SKUs and put them all on a menu that managers could select some combination of products they expected to win. We then integrated that with sales data that came in on a day by day basis and gave a custom aggregation of the product mix.
So instead of saying 1/2 Barbie 1/3 Hot Wheels or something like that, we got into a level of granularity for which they had never been able to plan. In time, leading syndicates began to be more obvious and instead of shouting matches about seniority and “I know this business like the back of my hand”, the senior management of Mattel had real data.
In every business in every industry, there are things that management does because they guess and there are things they do because they know. You would be shocked and amazed at how much guessing there is. And even where there is good guessing, you would be stunned to find how how much resistance there is against executive decision-making, not to mention sheer inability to distinguish rumor from actual policy (if and when policy is clear). Making business data-driven is a very big deal and most of the time, very difficult to accomplish. Even companies with the right data at the right time delivered to the right people can fail to execute.
If every business were data-driven, only few would fail. And of course if anyone could do it, we’d all be rich. You should never assume that management knows everything it needs to know about the operations of their businesses. They assume. They suppose. They engage in wishful thinking and they hope. Parents of teens understand this.
PS. You would think that when the company hired me as an expensive consultant, they’d have all of their data ready for my development work. Ha! It took a while for me to get access to the actual database with the product names at Mattel. In the meantime I had to make up names, thus Crackhead Barbie.
Comments