Death by proxy

Artificial Intelligence Cable Operators (MSOs) Decision intelligence Outcomes Telecommunications

“This year, our priority is customer experience.  Everything we do must connect to that.”

“We’ll upgrade the network in a neighborhood when the bandwidth utilization exceeds 80%.”

“I’m going to get rich, then I’ll be happy.”

Good ideas, on the surface.  Problem is: they’re often wrong.

And they share a common thought pattern: the use of a proxy—a substitute—for what we actually care about.

Think about it: if all that mattered was happy customers, you’d just send them each a thousand dollars every month and be done with it.   Similarly, if you’re a broadband provider that ignores factors like competition and speed of growth of network usage, then half the time you’ll be upgrading networks you didn’t have to, and the other half of the time leaving customers with slow Netflix.   Not a good use of capital.

Finally, as it turns out, money often buys misery. As Seligman and your grandma are delighted to explain.

The thing about proxies is that they’re easy to measure.  So in a simple and steady world, they can make decent shortcuts.

But sorry Toto, it’s no longer Kansas.  Today, the proxies are killing us. Stuff and money isn’t happiness, and we are wasting time and effort on business initiatives we don’t need in a hypercompetitive world.

We need one more link to our think. The good news: proxy busting is getting easier.  Why:

  1. An explosion of data that, along with machine learning/AI, can bridge the gap from proxies to real outcomes with a level of ease and power that was impossible just a few years ago.  And as new data comes in you can get better and better with that link.
  2. (yeah, you know I was going here, because I always do): we can use a decision model to understand, track, and analyze deadly proxy thinking.

Here’s the pattern:

Happy customers everywhere does not equal business success.  It’s an oversimplification that works sometimes, but not others.  We need “the missing link” that gets from the proxy to the true outcome.

(In DI speak, friends, this is the error of confusing intermediates and outcomes.  In our consulting work, we tease these apart by asking “why” if we think we’re hearing a proxy.  The answer moves us rightwards on the graph).

Please send me a list of the proxies in your world (they might make it into the book). Bonus points if they led you astray.  Super bonus if you overcame them.