The programmer's advantage marketers wish to have

For the last couple of months I’ve been coding–still am–the software of my new company PaperKup. As you can imagine it can be frustrated from time to time, and sometimes the worst feelings are when you run your code and it doesn’t work. An error pops up.
It’s weird the day everything runs smooth. But that doesn’t mean it’s a bad situation–it might be good. You can know right away what works and what doesn’t and make a guess of why doesn’t work. Usually you just have to make a search on StackOverflow and you’re done. There are hundred of ways to get a step-by-step guide, and if you follow them thoroughly, you get the desired outcome.
However, there are also times when you don’t know where the problem comes from, so you have to zoom-out. Not a big deal… You can go to a log file (where there’s detailed information of what’s going on), get a sense of what could be wrong, zoom-in again, find it and fix it.
In both cases, unless you’re creating something that hasn’t been done before, there’s almost always an answer to your problem.
But that’s quite not the case with other fields like marketing.
In marketing if the core of your code (name it campaign i.e.) is wrong, you’re still gonna see a version of your software. Which puts you on the hook. There are no errors popping out.
Therefore, you don’t only don’t get to see those errors a priori, you’ll just have a posteriori failure without acknowledging it. And by the way, there’s no log file where you can check what went wrong. You can get a bunch of data, with tons of metrics, but the conclusion is up to you.
Marketing is complicated, but considering these facts, you can set yourself for success following this two-step recipe (it’s not magical, but it helps more than not):

  1. Assume you’re wrong. I know, I know… you’ve heard that a hundred times, but now, for real, assume you’re wrong. Most people are anyway. Check the reasoning process in your guesses (I mean, hypothesis). Dive deep and find the flaws. In the end when you go to the core of the topic you’ll see that everything starts with an assumption of the world, which has to align with the market’s one.
  2. Measure things by final results, not things on their way to results. Measuring tweet conversions, not a big deal. Measuring exactly where the result comes from, that’s what you want. I see this over and over in the influencer industry where results are measured by how many views a video has. And that’s where you start adding errors to a log file you can’t check.

This is just a quick example, but there’s a bigger lesson to learn in here.
When we talk about social sciences the challenge is being able to learn to see and find errors. Unlike programming, social sciences don’t have flawless step-by-step guides. But if you’re able to find and fix errors, you become indispensable.