Time to Beef Up Your Algorithms

I had a good conversation with a smart guy in Marin County over the weekend about the next normal. One thing he said seemed to be aimed at the heart of CRM. This impressed me because Kevin is not a CRM guy, though he started and successfully ran a couple of businesses for decades.

Kevin’s contention is that the virus is removing a lot of spontaneity from society. After all, we’re a culture founded by people who picked up stakes and moved west — and sudden moves are something we know.

Before the virus, we shopped impulsively, made decisions instinctively, and acted spontaneously; long-term planning was not the strong suit of many citizens. However, when we wanted to buckle down, we could go to the moon. Daniel Kahneman, one of the fathers of behavioral economics, totally gets this and explains it in his books.

The Role of CRM

So, what about CRM in all this? Its main approaches currently involve capturing customer data in support of decisioning tools that do things like suggest the next best product or idea to support deal-making. But in the near present (a term my friend Steve G. likes), our algorithms, which are basically tuned to accuracy, might also need to be more tuned to precision.

What’s the difference?

Accuracy means getting the right answer, while precision is about getting the right answer time after time. In a CRM situation, an algorithm that gets the right answer is accurate and nice (we got the deal!), but if the business process supported by CRM has a lot of returns or unhappy customers, it’s not very precise.

The cost of imprecision is returns, which have a cost to bear — or lost customers, which have another.

So now here we are, mostly at home and just beginning the process of reopening. We’ve gotten by on a steady diet of Amazon deliveries, food drop-offs, and curbside service. The really brave have also done self-haircuts. Most of that has gone rather well (haircuts perhaps excluded), though delivery costs are high, and some things are just not returnable.

But imagine that the next normal features a lot more transactions through the screen. We’ve only been at this for a little while, so most people have not let their secret shoppers out. This means we haven’t really tested the limits of shopping for clothing. Getting things that don’t fit properly and require assistance to get right is a problem.

Better Algorithms for Fewer Returns

There are a few sites that already use the cameras on our phones to take our measurements and produce bespoke things that just have to be perfect (MTailor is one example). I suspect we’re going to see many more sites like this, and not just for clothing. At issue, the specialty retailer is going to become a software vendor, or the few that exist are going to be absorbed into bigger CRM solutions.

This in turn might lead to an increasing number of specialty processes mediated by even better algorithms — all in the service of making online shopping “return-proof.” That’s where spontaneity comes in. The time might be approaching when we don’t shop with the expectation that half of what we order can go back once we’ve tried it on or examined it with some accessories we already have.

I suspect this will happen most with non-luxury items that don’t cost a lot and therefore don’t have a big profit margin to support the overhead of the occasional return. Vendors of luxury items will continue to offer that kind of service because service is the heart of luxury.

But for the quotidian purchase, I’m betting we see tighter algorithms and more varied data collection to support them. Who knows? We might even see a personal phone app that captures and secures all of the data we really hate to share — like the size of our thighs, our weight, or our most recent optical prescription — the things we share on a need-to-know basis before we are allowed to make a purchase that can’t be returned.

We’ll have the virus to thank, and in 20 years, all of this will be great fodder for a cocktail party.