Trade Secret: The Right Response To Engagement-based Filtering
Less than two years ago, we were still debating whether engagement -- subscribers’ patterns of interaction with a given sender’s messages -- would really affect email deliverability. We’re past that now. It’s happening, and not just at Gmail. Yahoo and Outlook are clearly making inbox placement decisions based on subscribers’ engagement histories. (By the way, if you didn’t hyperventilate over the introduction of Gmail tabs this summer, the new Yahoo tabbed inbox gives you another chance. It looks great, it offers valuable insight into how Yahoo classifies and handles mail -- and like Gmail, it’s not going to kill commercial email, either.)
What makes engagement-based filtering a unique challenge to senders is that it varies so widely from one mailbox provider to another. Fixing your email program to gain deliverability at one ISP may not help you with others. But the good news is that however many different factors come into play in engagement filtering, it’s not a development designed to punish commercial senders. It’s not about them at all; it’s about subscribers. Engagement-based inbox placement is solely about using data from prior inbox interactions to make determinations on whose mail subscribers want -- and whose they don’t.
So the secret is to analyze your engagement as well as other senders’ engagement at specific mailbox providers, identify what’s working -- and not working -- and adapt your program using the same types of data that now drive inbox placement decisions.
We’re seeing three basic “flavors” of engagement-based filtering, so these may help you assess where you can improve your segmentation and mailing strategy.
First, there’s global filtering based on how engaged your list is with your messages. Subject line analysis, reactivation campaigns, and content tests can point to ways to increase subscriber interactivity with your messages. Simply cutting your list may not be the best course here.
Second, there’s global filtering based on how engaged your list is with all messages, not just yours. If a big part of your list is generally unengaged -- dead for all senders -- then cutting nonperforming segments makes more sense.
Third, there’s individual-level filtering, based on actual subscribers’ interactions with your messages. In these cases, customer segmentation can help identify which approaches resonate with specific personas, and highly customized email programs can tailor cadence and content to match individual tastes. Although clearly not a secret, this is the real key to beating engagement-based filtering: Use data to understand your audience and adapt your messaging to their preferences. By reviewing all three flavors of engagement, you may find that you are far from having a problem. If this is the case, the best advice may be to simply mail more (assuming you don’t have other reputation problems that could be exacerbated by mailing more, particularly issues with spam traps).
This isn’t merely theoretical advice. Senders are using subscriber data analysis and recommendations derived from it to change their mailing practices, especially for Gmail users. Most are seeing 10% to 15% improvements in inbox placement rate (IPR) as a direct result of engagement-based shifts. In a recent case, one of the best-known brands in the world studied the engagement similarities between international users whose messages bulked, hoping to diagnose a deliverability challenge with Gmail. Recommendations based on that study produced a 100% IPR increase, providing a long-term solution to a longstanding problem.
So that’s the secret: Mail better. Use the data that mailbox providers analyze. Study subscriber engagement, send more of what active users read, and less of what they don’t. As true individual filtering chips away at IPR, taking this approach can reverse the trend, get more mail into the inbox, get your messages into the hands of users that want them -- active, engaged consumers with a greater likelihood of buying from you -- and help your company make more money.