Marketing data. It’s a $20 billion industry composed of hundreds of companies, most of which gather, shape and sell consumer targeting data to brands and agencies. Very few of those companies, however, take steps to ensure or disclose any level of quality. There’s no methodological disclosure, no quality standard and no way to know how the flawed data may be negatively affecting campaign results. While it’s certainly available in abundance, it isn’t always solid, and it’s almost never proven.
In other industries, scientists reveal details and methodologies in academic journals, for peers to review and critique. That’s not the case with marketing data science. Why is that?
Jake Moskowitz from the Emodo Institute argues it’s less about our ability to understand what’s behind the data and more about companies’ efforts to veil data science, and he gives us six reasons why.
It’s only natural that if vendors develop a methodology that fuels their business, they’ll want to keep it under wraps. If it gets out there, someone can steal it, and when it’s the essence of your business, that could be risky.
- Sales Cycle
This is a business, after all, so closing the sale is pretty much the top priority. The deeper you get into the nitty gritty details of the methodology, the more questions that are likely to come up, and that can slow the sale down.
If companies don’t tell anyone how their methodology works, we’ll never know whether they actually follow it or not. But once they put it out there, suddenly they’re more accountable to follow it. Cutting corners is more difficult because the industry can hold you to the rigor you’ve previously claimed.
Vendors are constantly working to improve their methodologies, to fix discovered mistakes and make them stronger. But marketers don’t think of a methodology that fluidly. Once it’s explained to them, they have their understanding and that’s that. It’s not something they check in on, nor is it easy for vendors to keep everyone up to date every time their methodology changes.
In general, every vendor will have strengths and weaknesses to their methodology, just as we all do with all things. If a company breaks down every piece of its methodology, it’s bound to uncover weaknesses that they’d rather not reveal.
When we as humans find gaps in information, like hearing about an idea that we don’t fully understand, it’s part of human psychology to assume the best and fill in the blanks with something positive. So if a vendor tells us they use AI or machine learning — a field that most people don’t know a lot about — without explaining it any further, many people will assume it’s very sophisticated, even if it isn’t. So it behooves companies to leave gaps and let marketers assume the best.
With so much of your success riding on the strength of the data you use, it’s important to lift the veil on data science and find out what you’re buying.