A Portrait Of The Editor As A Data Analyst

I switched to digital publishing after five years of working in book publishing. And digital content is what I have been working in for the past three years. This article is meant to highlight a few key things that I had to learn (and learn fast!) in order to make my digital publishing career a successful one.

While working as a book editor, I’d had targets, obviously — there was a certain volume of sales (and therefore profit margin) that I needed to assure for every book I commissioned and/ or edited. But the challenge as a digital content editor was completely different: the target was month-on-month percentage growth in Monthly Active Users (MAUs) numbers. As a content startup, demonstrable growth in user numbers was crucial for fundraising; as well, since our revenue model was native advertising, the number of users was what we were selling to clients to generate business.

In the first couple of months I focused on setting up editorial processes and workflow to make the process of content production as smooth as possible. I figured that if the team got more efficient at ideating and writing, we’d be able to publish more stories. And with more stories would come more page views and users. It didn’t work out like that. We published more stories, but our page views and users flatlined. I needed to figure out how to publish better, not just publish more. And understanding the granular data that you get on a digital content platform was key to that. I needed to radically change the way I used data to inform my editorial strategy.

Here’s what I learnt…

Lesson 1: Embrace the beast that is your analytics dashboard

This was the first thing that I had to do. I had been shying away from exploring anything other than the ‘Overview’ tab on Google Analytics simply because it seemed, well, complicated. That was stupid. Whatever analytics tool you have access to, take the time to familiarise yourself. If looking at it by yourself isn’t helping, get someone else who has been using the tool to show you.

You can even use online tutorials to make your life easier. Google, for instance, provides online courses in using Google Analytics. You might also want to bookmark this very comprehensive article from Moz — the section titled ‘How to view Google Analytics data’ is the one you’ll want to keep going back to again and again.

Don’t just look at your topline numbers such as page views and users. There are other signals that your users are sending to you: with bounce rate, device information, referrers, session duration and so much more. Learn to read those numbers and see what they are telling you.

Lesson 2: Segment your traffic in order to understand what’s driving your overall numbers — and tweak your content strategy accordingly

One of the first things that I started with was device information. As it happened, 90% of our overall page views came from people on mobile devices — and it meant that I had to make sure that the majority of our stories were mobile-friendly. Here’s one way that data really made a difference to our editorial strategy. We were publishing a lot of lists with GIFs after every point because popular wisdom dictated that people wanted to read ‘visual’ stories rather than boring old chunks of text. However, the fact that people were primarily reading on mobile — on 3G (or even perhaps 2G back in 2014) — meant that those stories were incredibly slow to load.

We made a simple change — we put an outer limit to the size of a GIF being uploaded to the publishing CMS (<1 MB), and we also decided to use images after every second point or so. Within the course of a fortnight, our bounce rate went down significantly. Simply put, people were probably feeling less frustrated about looking at a blank page for several seconds while the story loaded, and therefore were much less inclined to not click on another story.

Lesson 3: Optimise for your top referrers

If you’re not running paid campaigns, chances are that the majority of your page views are coming from either social channels or through search. Especially if you’re doing non-news content. (Of course, you also have a certain percentage of direct views from people arriving directly to your homepage, but this depends largely on how well established your brand is.) Keep looking at your referring domains and optimise your content for them.

If the bulk of your traffic is from search, then obviously SEO best practices is something you need to quickly get on board with. If the bulk of your traffic is from a social channel such as Facebook, you need to embrace the best practices for that — right from titles that sound more ‘shareable’ to featured images not getting cropped awkwardly because they’re not optimally resized for your referrer.

Optimising for your top referrer is a long game — keep an eye on discernable shifts in referral percentages so that you’re always ready to change tack on your editorial strategy as needed. This doesn’t just apply to the ‘structural’ aspects of your content (story format, image sizes, title length, etc.) but also the nature of your content itself. Growth in search-dominant traffic comes from producing more need-based content (your articles answer questions that people have about things), whereas social-dominant traffic will grow when your content is more engagement-based (your stories are such that your readers will want to share them with their friends). Even your ideation strategy needs to keep your top referrer in mind!

Lesson 4: Write for your reader!

As a book editor, much of the projections I made about book sales were speculative. For instance, if I were publishing a thriller about a bank robbery, I would predict that the bulk of readers for it would be age 25–35, urban, male. There was no way for me to verify post-publishing that I was right. As a digital content editor, you don’t have that problem. Demographic data (especially if your users are signed in) is at your fingertips. If 70% of your readers are age 18–24, urban, female, produce more content that is likely to appeal to those readers.

If you’re looking to acquire a new demographic, depending on how aggressive that acquisition target is, strategise what percentage of your content you’re still going to be serving to your current user — not planning this out might mean a huge dip in existing user numbers while you’re still trying to figure out what’s going to make your new group of users come to you over and again. Retention versus acquisition is a fine balance — you have to be determined to ace it!

Lesson 5: Spend time on session details

Here’s something I discovered when I started looking at session details. Although our list-form stories had higher page views on average, session duration and page views per session were higher for longer-form stories than lists. Which meant that a single reader for a longer-form story was giving me more page views than I had anticipated — and was therefore a great candidate for becoming a returning user in the long run. I had to quickly re-jig our content balance in order to incorporate more longer-form stories into the editorial calendar — even if the impact wasn’t obvious before some time had passed!

Another fact that looking at session details revealed was that while at the beginning we had traffic peaks pretty much whenever a new story was published, over time time that pattern had changed and we were seeing traffic peaks at two specific times during the day, no matter when new stories were published. This led to a major shift in publishing strategy for us: we held back stories during other times of the day and started pushing out more stories during those ‘peak times’. Result? Even without going up on the number of stories being published, we started getting more page views — simply by surfacing content at times when readers were more likely to be reading!

Lesson 6: Track all your experiments

Any dynamic content strategy needs a certain degree of experimentation — you don’t want to lose out on the scope for innovation because you’re too focused on replicating success stories. The challenge in a data-driven editorial setup is that it can be unnerving to ‘try something new’ without a data bank to back up the decision. That shouldn’t stop you, though. Simply make sure to track those experiments as you execute them, so that you know which ones to pursue and which ones to discard.

In end-2014/ early 2015, as growth in video views started surging everywhere, we started publishing stories with embedded video links to see if that was a direction in which we could grow. We added a custom dimension for video as a separate category to be tracked by Google Analytics to easily generate reports for analysis. The experiment proved to be a success and led to the formation of a team of writers whose primary focus was to curate videos that were trending around the Internet and write features on them. Not only that, this data also proved crucial when it came to the decision to diversify content for the company, with in-house video production alongside text.

Lesson 7: Turn your whole team into a data-lovin’ one!

When I started out, we used to publish 100 stories a month; today we publish about 500. I had a four-person team three years ago; today there are twenty people I manage directly and indirectly. Basically, the more your content grows, the more stuff you will have to do through the day. So unless you want to become the bottleneck for decision-making — for success or the lack thereof — train other people in your team to try to understand data.

It’s a myth that ‘creative’ folks like writers don’t like to look at numbers. People who become writers do so because they want to be read — and if you’re able to show them how their readership can grow by them investing just a little part of their day into going through analytics, trust me, they’ll be be happy to do it! You might just be pleasantly surprised by the insights that they draw from the same data set that you’re looking at, things that you have missed because your eyes are tired, or maybe your brain is — I know I have been. And it has only helped us improve our editorial strategy over time. :-)

In summary…

There are factors that affect digital publishing that you can never fully control — a trend that dies sooner than expected; algorithm changes implemented by social platforms that affect the visibility of publisher content; a day-long malware attack that results in your web pages being de-indexed by search engines for weeks; a competitor popping up suddenly and cannibalising audience interest… There’s a lot of stuff that you, as an editor, simply cannot help. But your data is always talking to you. You just need to listen very carefully. That’s how you go from being an editor to a content strategist.

For the editor who’s just starting out in the digital space, I cannot emphasise enough the importance of the phrase ‘data-driven decision-making’. In the 33 months since I’ve made that my life mantra, we’ve seen 1,240% increase in the number of monthly active users and 1,467% increase in total monthly page views. And that’s something!

If you found this interesting or useful, do click on the❤️️ below. And if you have any comments or thoughts, I’d love to hear from you!

Also, a huge thank-you to Raoul Boström for reading an early draft and giving me tons of helpful feedback. :-)