The Trouble With Watson

I recently attended the Digital Summit in Philadelphia.

One of the booths that drew my attention while walking around was for IBM’s Watson. I approached it entirely because of a case study we did for class. I mentioned how incredible the technology was and how I was excited to learn just how it managed to win jeopardy.

From what I understood, Watson was being used in some advanced businesses. Healthcare, mainly, was what I knew of. I’m in email marketing, so how could it apply? I found out quickly that Watson is actually capable of being used in many more industries and channels than I originally anticipated.

I sat through the Watson presentation later that day after finding out just how it can be applied. But when I walked away, I had one question:

Is Watson worth it?

The trouble with Watson is that it easily ends up as one of three points in a triple constraint. Unless you’re a company with a lot of money (or a major healthcare provider), you end up with a classic triple constraint of timing, scope, and cost.

Watson is not cheap. While this technology is accessible, for smaller companies it would be a huge dip into the budget to use it. However, by bringing on Watson, difficult data questions can be answered in seconds.

In marketing, the scope of a project usually requires finding out a lot of information on your customers in order to better personalize for them. Sifting through the data in order to determine who should be marketed to and how is a time consuming process that usually leaves marketers wanting to pull out their hair.

So why wouldn’t you want to use technology to give you that data instantly? It means you could meet all your project requirements in half the time (or less) it would take to pull the data yourself. After all, anything you can learn along the way can also be understood by Watson, right?

Well, yes and no. Sifting through and organizing data can lead to a lot of insights you might not have seen before. And while Watson may have the answers for you, that doesn’t mean it necessarily knows the questions. Sifting through the data is a great way to come up with the questions you should ask in the first place.

And unless you have the budget for it, sometimes this painful manual process is actually worthwhile. You’ll end up with more questions than answers, but there are many cases when the questions are more important for your marketing strategies.

If your company doesn’t have to deal with the triple constraint, then absolutely you should put Watson online. But if you’re dealing with the aspect of budget, take a step back and evaluate. Sometimes, the answer to everything isn’t found through technology.