Your Startup Can Be a Finely Tuned Machine

I'm an engineer and one thing they teach you with any engineering education is to strive for building systems with deterministic behaviors.  As a technical startup CEO, I’ve had to figure out how to apply that principle to running a company. The challenge is that startups are inherently non-deterministic.  So how do you grapple with that?

About a year ago, I read Andy Grove’s book “High Output Management” and it reinforced concepts that I inherently knew from my engineering education, but had a hard time putting into practice as a startup CEO/manager.  I recently re-read it and made my whole team read it.  I thought it warranted a post.  I have to caveat and say that I don't subscribe to everything in this book.  Some of it is dated and feels very "big company", so take what works for you and run with it.  Ignore the stuff that feels dated.  But the chapters on thinking about management as a set of black boxes have been highly valuable to me.

The basic concept is to think about your company as a set of black boxes. You put the right inputs into the black box, build tight processes, monitor properly and you will get the right results out the other side.  If you're an engineer, this paradigm makes complete sense. 

With a technology company there are so many variables that at times it becomes overwhelming. You have design, product, sales, marketing, customer service, account management, employee happiness and morale, capitalization requirements, financial management, etc… Inside each of those disciplines there are even more variables to deal with. When you’re trying to manage all of this in the absence of a framework it becomes a daunting web of intertwined variables.

I always talk with my team about how it feels like those old cartoons where the character would come across a leaking dam that was about to burst. The character would inevitably stick their finger in the hole to stop the dam from bursting, only to see their effort result in another leak bursting springing up. Whenever you try and tackle one variable, you end up causing issues in other areas. It’s an endless game of playing chase your tail.

When I finally started getting a handle on things at CarWoo! was when I started thinking about my organization as a system of black boxes. Each black box needs to have expected results based on solid objectives. For example, sales is a relatively straightforward black box.  Revenue is your goal and your expected result based on your inputs. Some of the inputs are: a product to sell, sales leads that can be converted and enough staff to sell to your target market.  You put those things and and on the other side of the black box, you expect revenue.

What happens inside of the black box is a series of processes, best practices, tools of the trade and KPI’s (quality measurements). In sales that looks like pipeline management, CRM tools, hiring procedures, team management, etc… The manager is responsible for making sure the black box is a finely tuned machine and is working at peak performance.  If you set your goals properly, have a great person running the black box that can define great processes and you don't put in spoiled inputs, you have a great chance of that black box performing.  (The chapters in Andy's book about the 3 minute egg are really great.)

This thought paradigm helped me realize that you can tune your startup. And in order to win, your startup needs to be a finely tuned machine. Each of the black boxes needs to be running at peak performance. The set of black boxes is the engine that drives your startup machine.

Once I started to grasp this concept, it helped me realize there would be NO WAY I could manage this complex set of systems myself. Startup founders have a tendency to want to control everything, but this, in practice, becomes absurd. It has become clear that my job is to make sure I have great people managing each black box and that I make sure the team knows what the outputs of the system need to be.

I highly recommend Andy Grove's book on management.  It was written in the 1980's but the concepts, when applied appropriately to your startup, can really help you get a handle on managing your business.

Discuss on Hacker News.


special thanks to @millsbaker for reviewing and editing... and also reading Andy's book when I asked him to :)