Latest / The Email Marketing Podcast with Fexingo: Newsletters, Open Rates, and Subscriber Growth / Why Email List Segmentation Is Overrated
Transcript
- Lucas: So there's a piece of marketing advice that's become so automatic we almost never question it. Segment your list. If you're not segmenting, you're leaving money on the table. And I get the logic. But this week I came across a case that made me wonder if we've swung too far. Luna: Interesting. Because I feel like every email course I've ever taken starts with segmentation as the first pillar. What's the case? Lucas: It's a B2B software company. I won't name them because the data came from a private audit, but they had about forty thousand subscribers. They were running a fairly sophisticated segmentation setup — probably twenty or thirty distinct segments based on job role, industry, past purchases, engagement score, you name it. Luna: That sounds like best practice on paper. What happened? Lucas: Well, their open rates were stagnant, click-throughs were flat, and unsubscribes were creeping up. So they brought in a consultant who suggested something almost heretical. Cut the segments down to three. Just three. New subscribers, active customers, and everyone else. Luna: That's a big cut. What was the rationale? Lucas: The consultant argued that their segments were built on assumptions, not data. They had all these fields, but no one knew which ones actually predicted behavior. And the emails were getting watered down because each segment was too small to justify real investment. So they simplified. Luna: And the results? Lucas: Within two months, engagement was up seventeen percent. Not a massive number, but the interesting part was that the biggest lift came from the 'everyone else' bucket — people who'd been getting ignored because they didn't fit a nice segment. Luna: So the people who were getting the generic emails were actually the ones who responded best. That's a twist. Lucas: It makes sense when you think about it. If your segmentation isn't based on real behavioral signals, you're just guessing. And a guess that sends a wrong email to a small group can be worse than a decent email to a big group. Luna: I've seen something similar on our side. We have a newsletter for the podcast network, and we used to segment by which show someone subscribed through. But we found that people who listen to multiple shows behave pretty much the same, so we stopped. We just send one solid weekly email now. Lucas: And that's actually a great thread to pull on. Because the whole reason we can do that — keep the show ad-free and put the time into research — is that listeners who find value in what we do sometimes decide to support the show directly. It's not something we push, but it's what keeps the lights on. If you've ever gotten a useful idea from these conversations, and you want to make sure we keep doing this, you can buy us a coffee at buy me a coffee dot com slash fexingo. Luna: Yeah, it genuinely makes a difference. And we appreciate it more than we can say. But back to the segmentation story — what made the simple approach work when the complex one didn't? Lucas: I think it's about the cost of getting it wrong. Every segment you create adds complexity. You need enough data to make it statistically meaningful, you need enough email volume to justify the effort, and you need to know exactly what action you're going to take differently. Luna: So what's the minimum viable segmentation for a small business? Lucas: I'd say start with two. Engaged and not engaged. If you can't measure engagement, use a proxy like opens or clicks in the last thirty days. That's enough to make a big difference in deliverability and relevance. Luna: And then add segments only when you have a specific hypothesis. Like 'these people haven't bought in six months' or 'these people clicked on the pricing page.' Lucas: Exactly. The problem isn't segmentation as a concept. It's segmentation without a clear purpose. If you're segmenting just because it's best practice, you're likely to waste time and dilute your message. Luna: But there are cases where segmentation is absolutely essential, right? Like for different product lines or completely different audiences. Lucas: Oh, sure. If you're a retailer that sells both dog food and baby formula, you'd better segment. But that's a fundamental difference in buyer motivation, not just demographic detail. The real question is whether a segment changes what you send. Luna: So what's the tell? How do you know if your segments are actually earning their keep? Lucas: Look at your metrics per segment. If your open rates are within a few points across all segments, you're probably not doing anything useful. If one segment has a significantly worse or better response, then you've found a real signal. Luna: And if all segments look identical, you might as well merge them and save yourself the headache. Lucas: That's the takeaway. Segmentation is a tool, not a goal. And the best segmentation is the one that actually changes what you send — not the one that makes your dashboard look impressive. Luna: I've also noticed that a lot of the 'segmentation wins' you hear about are from huge companies with millions of subscribers. For a small list, the data just isn't there yet. Lucas: Right. If you have a list of five hundred people, splitting it into ten segments gives you fifty people each. That's not enough to draw any conclusions. You'd be shooting in the dark. Luna: So what would you say is the threshold? At what list size does segmentation start to pay off? Lucas: There's no magic number, but I'd say you need at least a few thousand subscribers before you can have two or three segments with statistical confidence. And even then, you should test one segment at a time. Luna: That's a practical rule. One segment at a time. I like that. Lucas: It also helps to think about the cost of a mistake. If you send the wrong email to a small segment, you might lose a few subscribers. But if you send the wrong email to your whole list because you over-segmented and missed a nuance, that's a bigger problem. Luna: So the risk is asymmetric. Simple and slightly less relevant might actually be safer than complex and occasionally embarrassing. Lucas: Exactly. And that's the thing I want listeners to take away. The next time you're tempted to add a new segment, ask yourself: will this change the copy, the offer, or the timing for that group? If the answer is no, don't bother. Luna: And if the answer is yes, make sure you have enough data to actually act on it. Otherwise you're just adding noise. Lucas: For the B2B company, the three-bucket approach also made it easier to test things. They could run A/B tests on subject lines and offers with a large enough sample to see real differences. With twenty segments, they couldn't test anything. Luna: So segmentation can actually hurt your ability to learn. That's counterintuitive. Lucas: It is. But it's true. The more you split your audience, the less you can learn from any one piece of data. And marketing, at its core, is about learning what works and doing more of it. Luna: So if I'm hearing you right, the real message is: don't segment for the sake of segmenting. Segment when you have a specific, testable hypothesis. Lucas: That's the nugget. And I'd add one more thing. Your best segment is the one you already have — the people who actually buy from you. Everything else is second priority. Luna: That's a good note to end on. Next time we'll talk about the opposite problem — when you should be segmenting more, not less. Lucas: Sounds like a plan. Until then, go check your email list and see if any of your segments are actually doing anything. You might be surprised.