ask more questions
In our push to be data-driven, do we risk becoming overly reliant on data? PHOTO: Jonathan Simcoe

Marketers are constantly being exhorted to use data and analytics to identify exactly what their audiences like, what they buy, what their motivations are and what they might spend money on in the future. 

Tracking and collecting all this information — so the argument goes — will help provide ever more targeted marketing. And, even though consumers might be creeped out by your ads following them around the web, that’s a small price to pay for personalization, the ultimate goal of the modern marketer. 

Meet the Tools of Data-Driven Marketing

That’s the party line, and today’s disciples of data-driven marketing don’t hesitate to use it as gospel to offer all kinds of tools for tracking and understanding customers and their behaviors. Those tools fall into three main categories:

  • Website analytics: These analytics track the sources of traffic to your website and how visitors navigate through your site while they’re there. You can even view a live feed of what a user’s mouse is doing on your pages with tools like Mouseflow.
  • Social media data: Social media helps you identify and connect with your customers, find out what they react to and even learn when they’re most likely to read your posts.
  • Predictive lead scoring tools: These tools provide insight into the prior purchasing behavior of certain customer profiles and predict purchasing patterns for how and where those customers are likely to spend in the future.

Are Your Data Goals Too Optimistic?

There are numerous tools out there which offer these capabilities, and many more possible ways to collect and analyze data on your customers. They will allow you to segment your audience even more narrowly, test what works and even calculate the ROI of your marketing efforts — universally considered the Holy Grail for marketers. 

Yet, as this research brief from MIT’s Initiative on the Digital Economy makes clear, companies that expect data and analytics to resolve all their marketing problems are probably being too optimistic. 

In the report, entitled, ‘The Limits of Big Data’s Competitive Edge,’ co-authors Anja Lambrecht and Catherine E. Tucker take a somewhat skeptical view of the big data revolution and point out that there is little evidence that merely possessing data offers a real means of creating value. For example, established companies normally have a huge wealth of data on their customers, yet small start-ups with no data but disruptive ideas can often trump even the biggest dataset. 

Caution, Data Limits Ahead

Their point is that access to all the data and analytics tools in the world will neither protect your company against marketing threats nor necessarily be able to intuit what your customers really want. 

What those caveats in mind, let’s explore some limits to existing data capabilities: 

Analytics can only analyze data that has already been collected

Even huge companies like Amazon can only collect information on the people who have used their services, so any analytical decisions they make are based on that data alone. This isn’t to say that marketers shouldn’t use their data, but they must be aware that there are countless reasons people will buy products, and only a tiny proportion of them are trackable. 

Data-driven decisions depend on a biased sample

Once again, the decisions you make based on data all depend on insights into a very limited, existing customer base. If the goal of your marketing is to reach new audiences, extrapolating from your data won’t necessarily yield useful information. 

Data can’t tell you why

Data can tell you what someone did but it can’t tell you why they did it. Did your customers buy a t-shirt because it was colorful, because of the season, because they had just spilled coffee down the one they were wearing or simply because they like it and bought on a whim? 

Consider Triangulation for More Credible Data

So, do these issues mean we scrap all attempts at using data in marketing?

The answer to this question is a resounding no. Data and analytics tools are clearly very valuable for marketers. But as we’ve seen, analytics will always depend on existing data, so analytical tools can only assist your marketing efforts in fairly restricted ways.  

Luckily, there’s an alternative — triangulation. This research method involves using at least three, and preferably more, datasets for research on a single issue. Long used by social scientists, the idea behind triangulation rests on the idea that one research method can only tell you part of the picture, whereas multiple perspectives will give you a more well-rounded view. 

Triangulation arose when social researchers studying, say, inner city gangs, realized that certain research methods had limitations when trying to understand their subjects. For instance, interviews were great ways of getting gangsters to talk about themselves, but the content of those interviews tended to skew toward positive self-assessments. Researchers needed alternative methods to validate their subjects’ actual behavior on the streets.  

Measure Thrice, Market Once

The key takeaway here is that data isn’t everything. If you want to really understand your customers, discover the actual reasons they behave in the way they do and provide them with products and services they really want, you will need qualitative information on your side, too. 

So marketers, you’ve still got a lot of work ahead of you, speaking to your customers to find out what they want and developing sophisticated strategies to get your messages across to them. Using data should certainly be part of that process but remember to triangulate your findings with at least two other sources of information before drawing your conclusions.