Editor’s note” Joe Procopio is the Chief Product Officer at Get Spiffy and the founder of teachingstartup.com. Joe has a long entrepreneurial history in the Triangle that includes Automated Insights, ExitEvent, and Intrepid Media. His columns appear on Tuesdays in WRAL TechWire.

RESEARCH TRIANGLE PARK – One of the most difficult skills for an entrepreneur to master is solid decision making. But decision making is the “Executive” part of Chief Executive Officer, and you’ve got to have it.

Listen, some skills you can fake, like leadership (not kidding, you can fake this all the way up to the end). Others you can farm out, like coding. Decision making? You’ve got to be good at it. No, you’ve got to be great at it. It’s one of the few metrics that directly correlate to startup success, where you can look back on your decisions and, to dumb it down completely, if you made nine good decisions out of 10, you’re probably at 90% of where you want to be.

Give or take a couple points.

The core of good decision-making is being honest with your criteria and honest with your results. We rarely tell the truth to ourselves and about ourselves at either end. No one likes to admit that fear and emotion play a large part in the process, no one likes to mitigate risk, and no one likes to dwell on their failures.

Joe Procopio

So that’s where we start. And once we’ve reached peak honesty, it all kind of comes down to pulling the trigger. There are philosophical cautionary tales here — “Damned if you do, damned if you don’t,” “Rock and a hard place,” “Should I stay or should I go,” and “If you choose not to decide, you still have made a choice.”

Yeah, those last two are the Clash and Rush but they’re probably copped from Ayn Rand or Camus or something very deep and intellectual.

Anyway, it’s not as convoluted as it sounds, and there are maybe three things to consider when you make a big decision.

Everything Is An Experiment

The first concept to get your head around is if you’re doing startup right, then everything you do is an experiment. It doesn’t matter how much the decision costs, how many people it impacts, or what it means for the future of your company — it’s only a test, whether you’re choosing a laptop or hiring a cofounder. It might work out, it might not. Nothing is guaranteed.

So loosen up.

But if all decisions are pointless, why make any in the first place? Because you have to. Because if you don’t, you’re going to stagnate and fail. Startup, despite what a lot of people will tell you, isn’t about surviving, it’s about winning. You can coast on your last success for quite a long time, but that time always ends with the doors being shuttered.

How much should I get paid to work at a startup? It’s more than a question of salary

When you make decisions, you should never be choosing between two options. Instead, you need to be choosing among several plans. Each plan should start with a choice, and then branch out like a tree to a number of different steps and outcomes. Each branch should be a hypothesis with its own definition of success and failure. You should continue to build off each branch and trim away the ones you abandon, and you should reassess each time you trim.

“This is no longer an option, what are my options now?”

What you’re doing has never been done before. There is no expert. There will be no wisdom of the crowd. It’s up to you to build the structure, ask the questions, and provide the answers. So draw up your blueprints in advance.

Let the Data Be Your Driver

Once you’ve got your experiments set up and your hypotheses ready to test, your “expert,” your “guide,” your “guru,” is going to be the numbers.

I don’t have to tell you what a minimum viable product is, but I can tell you that while a lot of people tout this strategy, few actually execute it. There are a whole bunch of reasons why they don’t — fear of putting out something that’s not ready, fear of customer backlash, fear of a reputation hit.

An MVP is not a coming soon page or a slide deck. It’s something you can put out there for real customers to use in real situations, and one that fails. If your MVP doesn’t break at the extremes, it’s not an MVP, and it’s probably got several bells and whistles built into it that aren’t going to be successful for you, no matter how much you or your customers love them.

What is an MVP? Let’s walk through one. So if I’m building Uber or Lyft, my MVP looks nothing like the app we know. Maybe all my app does is log you in, find your location, and present you with a nice big button to request a ride. Once that happens, getting you where you need to go can be a manual, duct-taped, non-repeatable, expensive mess.

Maybe you’ve got someone watching the back end waiting for a ride request, maybe you’ve got friends with cars running, maybe you’re using only cash for the transaction — it doesn’t matter. The point is that at each step in the process, you’re collecting data like how important is each step, how expensive is it, which parts are expensive, where do things go wrong. The whole thing might come crashing down, but you’re learning, and you’ve got enough backup plans not just to survive, but to win.

Once you’ve got data to support what you’re building, the decisions become easier. If you can save X or generate Y with a feature that costs Z to build and perfect, then that’s where you go.

Fail Often and Fail Quickly

Everyone says fail often and fail quickly. I believe in that, to a point. It’s a great way to talk about experimentation in a quick sound bite that makes the people who aren’t living it seem savvy, but it also scares the people who are living it.

Let’s break it down.

Fail often. You should always be willing to try something. If you’re experimenting, the bridges you burn are bridges that weren’t going to work for you anyway. You need to fight hard to prove your hypothesis, but at the end you’re going to break it down and rebuild it whether it succeeds or fails. No decision is either 100% right or wrong, you just need to be honest about how right or wrong you were.

If you’re looking at the decision as a road to take, you need to have routes off that road that you can use to go around the roadblocks. I change my ideas and plans probably a dozen times during execution. And honestly, if the whole plan proves out to be nonsense, I will abandon it along the way. Even though I’m giving the plan all I’ve got, I’m never married to a single idea.

The second part, fail quickly, is relative.

I’ve always said that you should take about six weeks to fully vet an idea before bailing on it. Some people think that’s crazy short, others think it’s way too long. It’s what I’m comfortable with, so do what’s right for you. Regardless of how long it takes, you need to be on the lookout for lifelines, little pings of success that might keep you going for months as long as you tweak and enhance and pivot off that original plan. Those lifelines can be good and bad. Ideas take time. But they also take resources, money, opportunity cost, bridge burning, and so on. So months, maybe. Years? Hell no.

Never fail in a vacuum. Talk to people you trust about what went wrong and why. Options that seemed like common sense to you at the time may have actually been insane, and the only way you’ll come to terms with that is to hear it until it knocks you off your position. We all have ego.

When you do realize your mistake, be proud of it. Lessons are always hard and expensive to learn, but regret doesn’t come from mistakes, it comes from action not taken.

Then the final step is don’t make the same mistake twice. 

This sounds like easy advice to give and take, and it is, until you run into a situation where you have the opportunity to pull that plan out of the bag again. All of us have that weakness towards insanity — “It didn’t work the last ten times I tried it, but this time is different.”

It’s rarely different. Not never different. If it were never different, it’d be a lot less tempting to repeat those mistakes. But no, sometimes a bad idea in the wrong situation is a great idea in a different situation. That’s rare, so always mentally mark down the chances that what didn’t work last time will work this time.

Instead, reset your experiment, get new data, make new plans, and pull the trigger when you’re ready.

Hey! If you found this post actionable or insightful, please consider signing up for my weekly newsletter at joeprocopio.com so you don’t miss any new posts. It’s short and to the point.