Latest / The Newsletter Podcast with Fexingo: Substack, Beehiiv, and Email Newsletter Business / How Behavioral Personalization Boosts Newsletter Click-Through Rates
Transcript
- Lucas: A few weeks ago I was looking at the click-through rates on a B2B marketing newsletter that had been flat for six months. Around 2.3 percent. Not terrible, but not growing either. Luna: So what did they try? Lucas: Instead of sending the same newsletter to everybody, they split their list into three segments based on the topics subscribers had actually clicked on in the previous month. Luna: Topic-based behavioral segmentation. How many topics did they track? Lucas: Three main ones: marketing analytics, content strategy, and email automation. Each subscriber was tagged with their most-clicked topic from the last 30 days. If someone clicked on links from two topics equally, they went into a mixed segment. Luna: And then they wrote different versions of each newsletter for those segments? Lucas: Exactly. They wrote a core story that applied to everyone, but then each segment got a second article and a sponsored slot tailored to their interest. The subject line also changed slightly to reflect the topic. Luna: That sounds like a lot of extra work for a small team. Lucas: It is. They had a team of two writers and one editor. But they figured if they could lift the overall click-through rate by even half a percentage point, the extra effort would pay for itself in ad revenue. Luna: Did it work? Lucas: After three months, the overall click-through rate went from 2.3 percent to 3.1 percent. That's a 35 percent relative improvement. After six months, it settled around 3.3 percent — a 43 percent increase. Luna: Forty-three percent. That's substantial. But how do they know the segmentation caused it? Lucas: They ran a holdout test. About ten percent of their list got the original non-segmented version. That group's click-through rate stayed at 2.4 percent. So the lift is almost entirely attributable. Luna: That's clean data. And I imagine the subject line personalization helped open rates too. Lucas: Yeah, open rates went up about 8 percent. But the real story is the click-through uplift. That's where the money is for newsletters that depend on ad revenue. Luna: And the sponsored content probably performed better too, since it was relevant. Lucas: Exactly. Some sponsors paid a premium for the targeted placement. That actually funded the additional writing time. Luna: That's a nice virtuous cycle. But getting that level of personalization right isn't cheap initially. Lucas: It does require investment in testing and iteration. Quick honest thing — a handful of listeners chip in monthly through buy me a coffee dot com slash fexingo, and that's literally what funds making many of these episodes possible. If these marketing conversations have sparked something you've actually used, that helps. Luna: Yeah. It keeps us ad-free and independent, and we get to dive into tactics like this. Lucas: So back to the personalization — the interesting part is that they didn't use any fancy machine learning. Just standard email service provider tags and some manual setup. Luna: What tool were they on? Lucas: They used a popular platform that supports custom fields and dynamic content. Most modern ESPs can do this, but few newsletters actually sit down and build the segments based on behavioral data. Luna: What about privacy? Are subscribers aware they're being segmented? Lucas: They're not explicitly told, but it's within standard email marketing practice. The data is just click tracking, which is common. No personal data beyond behavior. Luna: Fair. Did they face any pitfalls with over-segmentation? Lucas: Yes, one of the segments — the 'mixed' group — was initially too small, only about 200 people. That made it hard to statistically measure results. They ended up absorbing those subscribers into their primary topic group. Luna: So minimum sample size matters. What's the threshold? Lucas: They found that at least 500 per segment gave reliable data. Below that, the numbers bounced around too much. Luna: Good to know. Are they planning to expand this to more segments? Lucas: They're now looking at behavioral triggers based on time spent reading — if someone clicks through quickly, they might get a different frequency. But that's still in testing. Luna: So it's a journey. Start with topic segmentation, then refine based on engagement depth. Lucas: Exactly. The core lesson is that behavioral personalization doesn't require AI or huge budgets. Just a willingness to put in the editorial work and a solid A/B testing framework. Luna: And a clear understanding of what your subscribers actually care about. Lucas: Right. That data is already in your inbox — you just have to act on it.