Models, Decision-Making Processes And Snake Plissken

Startups are wonderful. Except, the goal of every startup CEO I’ve ever met is to escape Startupville as fast as possible. So, in the last 5-7 years, it has been important for the residents of Startupville to create escape plans. When some of those plans proved to be fractionally more successful than others, they moved from plans to models. Of course, in Startupville, billions of data points were collected to “prove” that a plan deserved to be a model. Once they were models, they became legend. Somewhere along the line, the residents of Startupville adopted these models and began to overlook their origins – as plans. I think this is an important distinction. In early stages, plans require their architects to create decision-making processes and evaluate the effectiveness of the decision-making process. If the decision-making process itself proved to generate a winning outcome more often than not, it was a good candidate to live a happy life as part of a model. In order to put a bow on a model, architects create measurement methodologies that guide people to use it correctly.

I liken it to a hobby of mine (that I’m not all that good at), woodworking. For example, Ian Ragsdale, VP Engineering of Boxer and I made him a pretty fantastic coffee table (pictured above). It took us 12-15 iterations on many parts of the coffee table before we developed an effective plan to create the production version of it. Our metrics were “squareness,” and “alignment tolerance” (and a subjective measure we called “do we like it”). We really liked building that coffee table, but it took forever. A year later Ian wanted a matching console table. It took us two weekends. We didn’t really think about it at the time, but cognitively, Ian and I had adopted a model for building his particular furniture style and so building a second one didn’t require planning. It required executing a model and measuring our results. (We had even built a ‘jig’ when we built the coffee table, which is a woodworker’s equivalent of a template that we were able to reuse on the second piece of furniture).

Something Ian and I didn’t do is question the decision-making processes that led us to a successful outcome on the first table. This actually proved to be a small problem. Turns out, the dimensions of a console table are different enough from a coffee table that our model and metrics didn’t do a great job of solving for a particular type of ‘bowing’ that the coffee table was not subject to. We’re perfectionists. We like the console table, but we love the coffee table.

I think the residents of Startupville often fall into the same trap. Startup CEOs (the decision making authority in Startupville) focus intently on other outcomes that are happening all over Startupville. Frequently, all they see are the metrics associated with those outcomes. They don’t have access to evaluate the decision-making processes that led to those outcomes. A problem is, what worked for someone else will not necessarily work for you – especially if you’ve done nothing to evaluate the decision making process they used to generate their outcome. Poor decision making processes can lead to good outcomes, good decision making processes can lead to poor outcomes. What’s a startup CEO to do?

I think it’s important, as a startup CEO, to not fall into the trap of using a model without making it a point to evaluate the effectiveness of your own decision making methodologies. Again, just because you got a desirable outcome doesn’t mean you’ve developed effective decision-making processes (guessing is a form of a decision making process, after all), and vice versa. Developing quality decision making processes as an individual and amongst your team is one of your critical responsibilities. Do not forsake it for “conversion rates,” “average order size,” or “viral coefficient.”

Finally, don’t be discouraged by an undesirable metric result. Take a step back and look at your decision making process. If you still believe it’s a quality decision-making process, then reuse it to generate a new idea. If you can’t put your finger on your decision making process (very common) or you no longer feel as though your decision-making process is viable, then you know what to do – stop using it to make decisions.