Windy

There are about a dozen sensors in a drone: It has a gyroscope, accelerometer, and barometer; a compass, LiDAR, GPS; and a handful more that even I don’t totally understand. 

While starting my first newsletter with the setup to a metaphor is sure to draw an eyeroll from Drew (I’ve been known to over- and misuse a metaphor or two before), I’m going to do it anyway. Because a drone—more specifically how a drone controls itself—is a pretty useful metaphor for ecommerce. (Have you ever seen a drone fly on autopilot? They literally follow you, carefully dodging branches and buildings, all while staying right on track despite a westerly gust. It’s insane!)

There’s been a lot of discussion in the last few weeks about tariffs. It is an interesting topic, to be sure, but it feels less interesting to me than many are making it out to be. The more interesting topic to me, at least, is how all this tariff chatter highlights one of the key constants in ecommerce: that variables are always at play and never constant. 

This is where my drone metaphor circles back.

Some know I studied electrical engineering as an undergrad. I learned about a concept called a Kalman Filter, an algorithm that recursively measures and estimates state. If you’ve ever used cruise control in your car, you’ve used one. It also happens to live in the drone’s flight controller firmware to help with stability, navigation and flight control.

Without going too deep into this metaphor (you might say I already flew way past that), a drone basically works in this way: 

  • The dozen or so sensors on a drone continuously provide data to the drone’s firmware about its altitude, velocity, orientation and more.

  • The flight controller firmware processes that data, using the Kalman Filter (or another algorithm like it), to combine all that data, determine which to “trust” the most, and figure out what corrections might need to be made

  • Based on those corrections, the firmware then sends commands to the motors, which, in turn, make any changes necessary to maintain, say, altitude and stability

So, if a drone is hovering and a gust of wind kicks up, the sensors will “catch” that, but it’s the algorithm that decides what to do with that information. 

Which brings me back to tariffs. Again: not tariffs specifically, but the fact that they’re a microcosm of the variability and uncertainty that’s hit ecommerce in the last several years.

There have been so many wind gusts, in fact, you might go so far as to call it windy: COVID, a shipping container crisis, stimulus checks, inflation, surging interest rates, iOS 14.5… the list goes on. I joke regularly that I am the least bit envious of the merchants I work with.

 

So, the point I'm honing in on, I think, isn’t so much how to respond to a wind gust (“there are new tariffs so I should raise or test raising my prices”), so much as what to do when it’s windy (“when my margin profile changes, I can respond accordingly because I understand my customer’s price elasticity”).

And this wind doesn’t always have to be the “Black Swan” type. 

What do you do when you’re presented with, say, a product that sells slowly (according to forecast) and you end up with excess inventory? Oh, and are you actively keeping a pulse on your conversion rates and CAC to know when something might be out-of-whack (and what to do about it in the moment)?

Of course you could just test whether you’re priced correctly, but price elasticity is only one signal, and changing prices is just one lever. And, like the conditions facing our metaphorical drone, the conditions facing brands are numerous and variable. Is the product poorly merchandised on your site? Did your competitors change their prices or positioning? Is an unseasonably warm fall affecting sales? Could a change to your welcome offer be the appropriate lever to pull? Or maybe a new landing page?

There are a lot of potential data points to capture in even this scenario. 

The only certainty these days is uncertainty, and ecommerce—more than any other business model—is well equipped to handle it. Retail, wholesale, brick-and-mortar stores, all of these have a great number of fixed constraints that prevent them from responding as quickly as an ecommerce business can. In fact, the best ecommerce operators I can think of operate very much like our drone, constantly detecting perturbations in the wind and adjusting the speed of each individual motor accordingly. 

That ability is most certainly a competitive advantage. These operators know at any given moment which levers are at their disposal and how hard to pull.  

When we zoom in on specific optimizations in ecommerce, it’s easy to obsess over, say, increasing conversion rate, or average order value, or lifetime value. It’s good for them to all go up in concert, obviously, but that rarely happens in a perfect environment, let alone an uncertain one.

So the advantage comes with being able to detect each change in the wind, whether at your back, in your face, or from the side. And knowing how to course correct. 

It’s the Kalman Filter—but for ecommerce. And you’re going to have a rougher time flying high if you’re not using it.