The Comfort Of Decisions with Data
This post is very short and reflects just my opinion. Do note that I am no veteran and do not have the kind of experience to have 100% valid opinions. Take all this with a grain of salt as I, myself, am learning. And with constant learning comes changes in opinion over time.
“In God we trust; all others bring data.” ― W. Edwards Deming
Ever been in the driving seat of a decision? I certainly have, and let me tell you, knowing that a decision can impact several future events and be a make-or-break deal is both empowering and uncomfortable. The dreaded “I told you so” looms over your head if the decision goes wrong.
This post is about how making data-based decisions has a sense of comfort in it.
Let me elaborate. When deciding, you want to optimize for success and minimize failure. I use the word “optimize” because in most situations you can never know the outcome with certainty.
But there is an often overlooked fact, even if the decision fails, we want to minimize “guilt” and “scapegoating”, and to minimize them, you need to be able to point a finger at something and say “See, my decision was based on this, and hence I had good enough reason to pursue it.” and that is where “comfort” lies, there’s backing for your decision and not just your gut feeling.
Sidenote: Minimizing guilt or accountability is the only reason management consultants exist, so corporate executives can point to their reports and say “See this report from XYZ Consultancy Firm predicted”. Now, as long as there isn’t malicious intent or conflict of interest (An AI provider sponsoring an AI report that shows AI will be a game-changer) behind that data, it’s all good.
Feel there is no data for your decision? Think deeper
We often come across situations where we think there is no data to back or make a decision on. All we need to do in such a situation is look slightly deeper, while an exact dataset on XYZ might not be available, other data that influences XYZ definitely will be.
For example, for an e-commerce company to decide whether a product will sell well, data on how many users purchased it before it has even launched is obviously not available, but data that tracks how many users visited the marketing page for the product and added it to their carts/wishlists might be available. This is an elementary and straightforward example that is naturally understandable to people.
Things get complicated when we talk about decisions that involve second-order effects, or data that “influences” a decision but isn’t a direct indicator of it. For instance, if you run a B2B SAAS business that generates a crucial sales document for your customers. Your software has other features but most users just use that feature as it’s the primary value-add. Now, you want to streamline that process and add a “Quick-document” feature but you want to assess whether it will be useful and whether quickly.
To judge that, you already have an indicator, the number of documents your customers generate. But a better indicator could be “How many customers come to the software, generate the documents and immediately leave after?”. Those customers would be the target for such a feature, where you double down on the utility and extract less revenue but higher volumes of customers by providing them just that feature and nothing more for future customers too.
What if the decision goes wrong?
Most decisions are oversold in their impact, both in your mind and by the people asking for a judgement.
Decisions can be reversed, such decisions are called Two-way doors. Chose the wrong library for your project? Just get a new one.
One-way doors are decisions that cannot be reversed or just take too much effort to roll back. Hired the wrong founder for your company? Gave a mission-critical project to the wrong person to get done on a tight deadline?
It is thus recommended that even with data, for one-way doors you take as much time as possible and combine data with your gut feeling to ensure you’re making the right choice.
That being said, most of the time, decisions that seem to be one-way doors are in fact two-way doors when thought of at a deeper level. Heck, a startup that has signed up hundreds or thousands of customers will find it very difficult to pivot, especially when its entire strategy was built around an initial idea. Yet, if the current model doesn’t make any money, a startup pivot remains one of the most common occurrences and is usually seen as a good thing if the startup is confident they’re pivoting to a more viable business model.
Talk to any fintech founder and they’ll tell you at least one instance where they have had to throw years of their work down the garbage dump and start from a blank slate because a regulatory change rendered their company invalid overnight (In fact, they have had to bake in and live with that possibility from the start). When dealing with people, people understand, as long as there’s no malice in your decisions.
Even if it is a true one-way door, life is long and you have ample time to recover from any failures from your decisions. Treat them as battle scars that show you have experienced burning your hand with something and making it out of there, better and wiser.
Gut Feelings
All the above being said, there are scenarios where the data might not line up to your expectation but you have conviction in your mind against it. This is often the “Gut Feeling” people talk about when all the data in the world can’t convince you for/against a decision.
In such cases, if the decision doesn’t have a destructive impact, it’s better to just go with your gut feeling. Contrary to popular belief, going with your gut is not a regret-minimization framework, because there’s a good chance that your gut feeling might be wrong and that might lead to regret.
That being said, several bets in the past that changed the course of history have been based on gut feelings. You just have to define the risk that you’re willing to take, and whether you’re comfortable with “I told you so” after something goes wrong because everyone will simply point to the heaps of data that you went against.
There’s comfort in data, but there’s conviction in gut feelings, decide what you want to go ahead with.