Please Help, I’m In An Abusive Relationship With My Marketing Data!

 In data analytics, Data-driven Marketing, Standard

I suppose when you’ve been in the digital marketing space long enough, frustration will inevitably take hold when working with your marketing data. I think I do a reasonably good job of brushing off the typical day-to-day stress inducers while trying to make sense of analytics, but one thing occurred to me recently: Some of the things that cause my left eye to twitch ever so slightly are just too frequent to let stand. Enough is enough, and I must therefore admit to you that I’m indeed in an abusive relationship with my data.

An abusive relationship is one that can be defined by how you are meant to feel by the alleged abuser. Do you often feel humiliated, your accomplishments belittled? Are you embarrassed for your friends and family to observe your interactions? I submit to you that marketing data has so often done these things to me that I had to reach out for help and identify this problem.

If you find yourself in a similar situation, please know that you’re not alone. I sympathize with you, my fellow marketers, and for this purpose I have provided the following top five reasons why marketing data should be treated carefully. Hopefully these insights will help you as much as they did for me to confront and alleviate the problem for a lasting, loving and healthy relationship with your data:

1.    Wishful causality

Many marketers look at unrelated correlations and assume causality. Is my spike in leads for this hour really a result of my tweet from an hour ago? To be fair, this is not the data’s fault; it’s actually your fault for mistaking happenstance with causation. Sometimes you just have to realize that if the data did in fact show you that that spike occurred precisely six minutes after your tweet went out, it doesn’t necessarily mean it had anything to do with it.

2.    Garbage in…

You must strive to understand the full lifecycle of your data before jumping to conclusions. For instance, if you’re looking at demographic data obtained by a third party, how was it generated? Was it based on government census data where the visitor lives, or based on an IP address lookup? Did you assume that the visitor told you their age or income level directly?

3.    Overbearing

For every piece of useful relevant data, there’s a hundred pieces of complete and utter nonsense that you shouldn’t be bothered with. Why must I know which model of Samsung phone my users have? Sure there are hypothetical use cases for such things, but you are much better served starting with the business question then working back to what data you will need to answer the question.

4.    Delayed reality

While it’s certainly true that you can make decisions rashly, you can also take too long to make them. With so many factors influencing the success or failure of online marketing efforts, can you reliably say that what worked last month will work this month? The notion of “right time” may be useful here.

5.    In one ear…

Some data is just plain misleading when sliced certain ways. If you’re not correctly correlating customers across devices, for instance, you may believe that a particular product is unpopular without getting the whole story. Upon further analysis, you may discover that the tablet experience is extremely popular, but the same product viewed on a desktop device is incredibly unpopular. Or in an aggregated view it could look like the product is doomed, but when separated you may find that the product page is just broken for non-mobile devices.

It is my hope that some of these abuses will now be recognized and addressed, and marketers everywhere will regain some dignity and demand to be treated with the respect we deserve.

Let me know if there are any other examples of abuse that you would like to share. Only together can we put an end to the unhealthy relationship with data many marketers have yet to face.

Enhanced by Zemanta
Recent Posts