Marketing Analytics: Three Common Data Pitfalls

When you’re as data-driven as I am, working at a data company is really satisfying because analytics are embedded in every department. In fact, when I started writing this piece, I reached out to my marketing colleagues about common pitfalls with data and they very quickly filled my inbox with responses.

In my , I talked about how marketing as a function is becoming more data-driven (even though PR specifically still has a ways to go). Although marketing departments are rapidly evolving, we’re all still learning best practices surrounding when data is an asset — and when it can be a liability. So I went straight to the source — some of the most data-driven marketers in the industry — and asked them where they see the biggest data mistakes and what they would do differently.

Problem One: The Silo Affliction

We have a tendency to look at the data we think is most relevant to what we’re doing. But data doesn’t live in just one department or unit. Monitoring and analyzing data in a silo restricts us from seeing the insights that could give us a better understanding of what’s happening holistically. For example, you’ll get a better read of your press coverage data if you look at it from the lens of your social engagement, demand gen, website traffic, etc.

And just one of these isn’t enough: The more data you can integrate into your own, the better your read and insights will be. For example, web traffic stats are often siloed. Pay-per-click (PPC) and search-engine-optimization (SEO) metrics are usually reported separately from email and social media. As a result, a lot of companies don’t know to what degree each channel affects the other.

Problem Two: Misattributing (Or Misunderstanding) Data

By integrating data, a web traffic manager and social media manager might come together to say we reduced our spend on social media ads this month and saw organic traffic decrease 10%. Cleaning, joining and blending data, and allowing people to see it side by side, makes it easier to spot correlations between channels. In volume, you can predict how a business decision will impact your customers’ experience with your brand across all channels.

One of the things I talked about in my last post is that sometimes using data in PR is tricky because it could potentially (or accidentally) disprove the impact of articles. Using siloed data could misattribute cause and effect, but PR isn’t the only place that can happen. For example, as my social media colleagues informed me, engagement rates across different platforms (LinkedIn, Twitter and Facebook) can actually mean different things because each platform’s ratios are very different between channels and industries (read more ).

To that end, reach and impressions can mean different things, and not fully understanding that could cause the same misattribution. SproutSocial wrote a great blog post about the difference between reach and impressions and why they’re often misinterpreted.


Article by channel:

Read more articles tagged: Marketing Analytics