In the mid 1950's a series of simple experiments were held with various groups of people to get a handle on the cognitive capabilities of people.
In a famous experiment, Baba Shiv randomly split groups of people into two groups:
- One group memorized two digits, say “17”; and
- The other group memorized seven digits, say “3784926”
Before each group reported their number, they walked down a hall where they were offered a reward for their efforts – either a healthy fruit cup, or a piece of chocolate cake. People who were to recall the seven digits were 50% more likely to choose the cake.
Why was that? It's because the mental effort required to remember those extra numbers induced them to rewarding themselves with the less healthy cake. Researchers call this effect “cognitive overload” and its unfortunate effects are well documented.
As companies expand and mature, rather than rationing or reducing load, they often pile on so many metrics, procedures, processes and activities that people lose the capacity and capability to do the right things.
In the mid-1950s, psychologist George Miller showed that people could hold roughly 7 numbers in short-term memory. This became known as Miller's Law during thousands of observations on the hazards of overtaxing our brains.
Miller's Law has wide-sweeping implications for potential flowcasting solutions and their ability to scale.
If you understand the paradigm of Flowcasting you'll quickly appreciate its power and benefits. The supply chain, completely driven by a store/sku forecast, entirely recalibrating based on what is and isn't selling, daily.
Many retailers have 10's or even 100's of millions of store/sku combinations. The ability of a solution to scale to these kinds of volumes is not an easy proposition.
To date, I've been invited to review a few of the major technology providers who beat their chests and proclaim, before my visit, “we've got a Flowcasting solution”. After my visit they realize what they've really got.
I'm not shitting on these companies, as they are fine organizations with a long and illustrious history of providing supply chain solutions. All the companies I've visited view Flowcasting as adding more to their existing solutions – essentially thinking that they can just extend the solution to the store and add lots more computing juice to crunch the numbers.
I think these companies, along with others attempting to build a Flowcasting solution, would be well advised to think about Miller's Law and try a fundamentally different approach.
You see scale is not a technical issue, but rather a people issue.
I'll let you in on a little secret that very few people really understand. Please keep this to yourself, if you can.
All retailers have essentially the same business processes.
What that means is that the mid-size retailer I'm currently working with has the same processes as a behemoth like Wal-Mart. They introduce and discontinue products, promote, add new stores, add DC's, change product flows and buy product from all over the world.
The issue with trying to take existing solutions and add to them is one of cognitive overload. For people. The system becomes so heavy and difficult to implement and manage that it collapses under its own weight.
It becomes too burdensome for people to use, implement and sustain. So what do they do? They go around it. They use spreadsheets and manual processes.
I've seen recent studies that highlight that 60%+ of companies don't use their planning solutions and rely on spreadsheets for planning and ordering. Even some of the largest and, theoretically, more advanced organizations.
Suppose a technology company said, “fuck it, we're going to do this differently”. “We're going to design our Flowcasting solution, virtually from scratch, to be easily implemented by small or mid-sized companies”.
The beautiful thing about an approach like this is that they'd quickly realize that smaller retailers would have no appetite, desire or capacity for unnecessary complication.
All they could handle and support would be what is required to get the job done.
Yet getting the job done in a small to mid-size retailer is the same, from a process perspective, as a large retailer. And the resulting solution to handle this would be clean, crisp, elegant, simple and, ironically….
…able to scale.