The five-step algorithm simplifies before it automates

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

1. 56:36

   What is The Algorithm? It's a formula that we use, or really a framework, I would say, that we used to create these breakthroughs. and to create innovations. It was a framework that we developed by doing postmortems and on a series of big mistakes that we'd made and asking ourselves, how do we not make this mistake again? And we came up with, over time, this five-step algorithm that would help us not repeat mistakes.

2. 57:06

   And the counterfactual was it would drive innovation. So it turned out to be a really powerful framework that's in use today, like in the spacex ipo in the first few slides there's a slide called the algorithm uh where they say hey like if you're going to invest in this company you need to know how this company runs and we run on this framework of the algorithm and it's what led to the advancements from raptor one to raptor two to raptor three of the engine and very much tesla's built that way too can you maybe double click on that walk me through the five it's all a journey in simplification uh and so you kind of get that with five steps for the First step of simplification, you want to remove all requirements that aren't real or dumb.

3. 57:47

   And so you question every requirement. And so that means things like if you have a 12-page loan doc, you question why you need every paragraph in that loan doc and why you can't get it down to one paragraph. So you eliminate as many requirements as you can is the first step. Now you've got a simpler set of requirements, and then you design a new process to deliver those. And you delete as many steps in the process as you can.

4. 58:12

   And the hack that I would use with my teams is I would say, map the process out, like literally put sticky notes on a wall for each step of the process, and then go circle the ones the customers pay you for. And it turns out in almost any business process, a super minority of those the customers pay you for. They don't pay you for documents. They don't pay you for communications. They don't pay you for HR benefits, all kinds of stuff.

5. 58:36

   They don't pay for that. And so you've got to figure out, okay, how would I deliver this in a way that is totally maximized around what the customer is paying for? That's step two. Step three is now you have a brand new process that has been very simplified. Now you need to start to run it and optimize it and polish it manually. And step four is you apply the great kind of revealer to that process, which is speed.

6. 59:04

   When you speed up a process, it shows all of its warts. It shows where it breaks. It shows its failure points. It shows quality points. And so you've got it. Speed allows you to, like, see all the failure points, and you can then scrub those out of the process. And then once you've got a process that's working at speed optimally, then you automate. And so the last step of the algorithm is you automate last.

7. 59:26

   And we had made mistakes, like, automating first. we'd made mistakes like just not questioning requirements and ended up in like multiple month loops on problems that wouldn't go away and so this whole framework allows you to not only solve problems but innovate um because you're you're eliminating requirements that your competitors think exist but they might not need to exist outside of manufacturing can you give me like a recent example of how you've used that to solve a problem?