Data-driven content marketing has been a cornerstone of RJMetrics’ marketing since the very early days. From Jake’s 2010 investigation into Twitter user engagement to more recent coverage of Meetup, Ello, and Etsy, we think that telling stories with data is simply more compelling.

We’re not alone in realizing this. Data journalism, nascent when we began, is now a large and growing trend, with folks like the New York Times, FiveThirtyEight, and Vox leading the way. There’s even a data journalism handbook.

Conducting original research, sourced from original data, is how we stand out from the cacophony of lowest-common-denominator content that is prevalent in the market today. It has become easy to create content, but it remains hard to do it well. Trust is critical to this strategy. You have to get the data right, and you have to have a history of doing so. More on this later.

We’ve had an impressive amount of data journalism success: these stories have had hundreds of thousands of views, tens of thousands of shares, and often do well on both Reddit and Hacker News. Data journalism has created more brand engagement for RJMetrics than any other marketing tactic.

In 2015, we’ve decided to double down on data journalism. We’re investing in people, processes, and technology focused on creating more original, data-driven content about online businesses.

What exactly are we doing? Specifically:

  • We recently hired a full-time data scientist to focus on data journalism. Yevgeniy Slutskiy, who has a PhD in Computational Neuroscience from Columbia, will be joining our team this coming Monday.
  • We built a data scraping infrastructure. With it, we’ll be able to gather data from the public web faster than ever.
  • We now have a full-time writer dedicated to data journalism. David Williams has 12 years of experience in marketing and is the author behind our 2015 Ecommerce Growth Benchmark Report.
  • We’ve implemented an analytical review process to ensure that every piece of data we publish has been audited by multiple sets of eyes on our team.

The last point on that list deserves a bit more discussion, and some credit where it’s due. We recently released our 2015 Ecommerce Growth Benchmark Report, which was downloaded over 2,400 times in its first week. One of those readers—Ben Sun from Primary Venture Partners—noticed what looked like an inconsistency in the data and was kind enough to bring it to our attention.

Upon investigation we realized, much to my personal horror (I had conducted the analysis!), that an analytical error had made its way into the final version. The problem had to do with the “Customer Lifetime Value by Quartile” chart on page 17; we overstated the difference between Q1 and the other quartiles. After correcting the initial error and taking a couple of deep breaths, we set about creating a process that would keep this from happening again.

We now analyze data in much the same way that software developers write code. We have development environments where we run whatever analysis we want without any guarantee of it being right. Once we’re convinced we have the right answer, we submit Github pull requests to our team members for review. Once a pull request has been approved, that analysis becomes fair game for writers to publish.

Honestly? I’m glad that this happened. I’d rather work out problems sooner rather than later, and this gave us a great opportunity to do just that. I’m very happy about the changes we’ve made and am more confident than ever in the data we publish.

We’ve re-launched our 2015 Ecommerce Growth Benchmark Report in the brand new web-based format that will become our primary mechanism for publishing all of our research moving forwards. We’re excited about moving into a richer, more interactive environment and are excited about continuing to play around with the format moving forwards. Check out the updated report here.

Data journalism is the centerpiece of our content marketing, and we’re only going to do more of it. Stay tuned for lots more research. Please drop us a line if you have suggestions for topics we should dig into.