“The Sell Sider” is a column written by the sell side of the digital media community.
Today's column is written by Christian Baesler, president at Bauer Xcel Media.
Publishers have a lot at stake when deciding to test Facebook Instant Articles.
Some skeptical publishers think of Instant Articles as a way for Facebook to commoditize the online content space more quickly and have decided to opt out. Others, including The Washington Post, have fully embraced Instant Articles, thinking of its success as a matter of audience reach rather than one that relies on specific pieces of content.
From experience, I think it’s important for publishers to take a scientific approach. To use a cooking analogy, testing risky bets such as Instant Articles should be approached more like baking than grilling a steak.
In other words, it’s not about “feel.” You need to follow the recipe and measure precisely. Any company using Instant Articles should strive for as much insight and control as possible.
Measure From The Top Down
I’ve talked to a lot of publishers that are not testing Instant Articles effectively. The value in Instant Articles compared to regular link posts needs to start with a proper A/B test. While publishers can’t post the same content in both versions at the same time, they should try to get as close as possible, by either posting on Instant Articles similar content or evergreen articles, such as recipes, that would appear a few days later as traditional link posts.
I also see that many publishers fail to account for everyone who sees the post, not just the people who see their ads on the articles themselves.
Publishers should look at the total addressable universe of Instant Articles compared to regular posts by calculating three numbers: reach, engagement and click-through. This way, there are multiple points that can be tested to increase their addressable audience.
The first step is to calculate the reach of the Instant Article post in the news feed, not the article itself. With this metric, you can adjust what you post and determine how that audience changes and even note what the Facebook algorithm might prefer.
Next, measure the engagement. Again, I’m referring to early stage, high-level engagement. Are people more likely to comment on or share Instant Articles than with a link post? This metric gives you the opportunity to test. Does changing the title or content lead to an increase in engagement?
The last step focuses on the click-through. Which type of news feed posts yield a bigger audience for your actual article? Again, this metric gives you an opportunity to test ways to improve from the beginning of your interaction with consumers. This converting audience is the one that you’ll be directly making money from.
Revenue Comes From Reach
Now that you have a sense of your total reach across both post types and an understanding of how to maximize traffic and engagement, you can start thinking about the revenue you are earning.
But don’t just look at people on your article page; calculate the RPM based on the impressions that the post generated on Facebook. You want to know how much revenue you earned based on impressions for those who saw the link post in the news feed, not just those who read the article and thus saw ads on your site. This shows the number of people who could be converted to readers.
Take into account this high-level reach and you can make a clean comparison between the two posts. Then you are empowered to measure the effect of changes from the top down, not just at the actual points of monetization.
In the limited universe of Instant Articles, ads and a few links to your website are the only two direct revenue calculations to be made, although recently Facebook did announce it will start allowing publishers to collect email addresses for newsletter signups on a limited basis.
On the other hand, a traditional link post includes a variety of revenue opportunities on a publisher’s own site, from ads on the current article to those on additional articles read, to conversion opportunities, including newsletter sign-ups or transactions, such as subscriptions or product sales. It’s important to look at the revenue per user for their entire visit, not just the CPM earned on the specific article in question.
Account For What You Can’t Do
Don’t forget that there are many issues with Facebook as a publisher partner that bleed over to Instant Articles. Making an accurate calculation of the lifetime value of a visitor to your site should take into account what you can’t do on Instant Articles posts.
Publishers really don’t get any in-depth audience measurement opportunities on Facebook. Facebook keeps their audiences to themselves, so creating a full picture of a viewer’s browsing history is limited by using Instant Articles instead of links, not to mention the lost data that could be used for future targeting.
Second, advertisers have limited options since slideshows, advertorials and other creative ways to increase revenue are not available on Instant Articles. In fact, for now Facebook only allows one ad every 350 words and only a couple of ad sizes with no rich media, which is bad for viewability and engagement. Most importantly, the close connection to the reader is lost. When a reader is on a publisher’s own site, it’s much easier to collect data, engage and generally build a long-term customer relationship.
For publishers that work with advertisers beyond branding to create opportunities for conversions – retail or otherwise – their own website is likely a better platform for even the most basic relationship-building efforts such as gathering email or registration information. A detailed measurement plan that takes these issues into account when calculating the comparison between your options will help you decide how Instant Articles fits into your strategy.