The pitfall of a standard data solution

Take a marketing campaign as an example. Suppose your last campaign is not giving you the results you had hoped for. You want to improve by making it more data-driven. You therefore ask your analytics team for help. They are happy to help and talk to you about your campaign. They ask you some questions to better understand what kind of campaign you had set up and then come up with a suggestion: they can optimize your selections with an algorithm. This way you can find the most promising leads with less effort.

This can be a good solution, but it is not always the best solution. In fact, data science can make your work a lot more efficient, but when this becomes the default solution, there is a chance that you will overlook something else. By zooming in too quickly on one part of the process, you might miss another aspect where much greater gains can be made with much less effort.

An end-to-end smart solution

Take the campaign example again: If the weakest link in your entire process is the selections, then an algorithm (provided the boundary conditions such as data selection are well thought out) would give you a lot of upside. However, if the selections are not properly aligned with the goal of the campaign, optimizing them is going to be of little use. This is why it is important to always look at the whole end-to-end process when you want to make something “smarter.” First figure out where the weakest link in this process is and try to improve it.

Sometimes that will still be an algorithm, but in many other cases it can be a simple Excel analysis, a dashboard or a good A/B test. The only way to find out is to step back and evaluate the whole process.

So what is a “smart solution” to us? It is finding the right solution to the problem at hand. No matter how simple or complex that solution is.