Latest / Marketing Automation with Fexingo: HubSpot, Marketo, and Automated Customer Journeys / How Marketing Automation Workflows Miss Customer Communication Preferences
Transcript
- Lucas: So there's this moment every marketer knows — you spend weeks building a perfect automated journey, you map every trigger, you write the copy, you set the timing — and then engagement flatlines. Not because the message was wrong. Because it went through the wrong pipe. Luna: The pipe meaning the channel — email, SMS, push notification, whatever. Lucas: Exactly. And the interesting thing is most marketing automation platforms are great at letting you set channel preferences once — during checkout or sign-up — but terrible at keeping those preferences alive over time. I've been looking at a case from a mid-size e-commerce brand that switched ESPs last year. After the migration, their email open rates dropped 14 percent in two months. They blamed the new platform, ran deliverability tests — nothing. Finally they looked at the preference data and realized that during the migration, 22 percent of their active customers had their channel preference reset to email. Those customers had originally opted for SMS or push. Luna: So the system effectively overwrote their choice. That's a data integrity problem disguised as an engagement problem. Lucas: Right. And I think the deeper issue is that preference management is still treated as a single checkbox — email, SMS, both — when in reality it's a multi-dimensional dataset. People have different preferences for different message types. I might want transactional emails but promotional SMS. Or weekly digests via push but daily alerts via email. Most workflows just don't support that granularity. Luna: And preferences change. If I signed up for SMS two years ago, I might now ignore text messages. But the workflow doesn't know that unless I explicitly update my settings. Lucas: That's the other piece — preference decay. There's some data from the Email Sender and Provider Coalition that suggests about 30 percent of consumer channel preferences shift within a year. Life changes — new job, new phone, new habits. But the typical marketing automation workflow checks preference once and then assumes it's static forever. Luna: I think a lot of marketers assume that if someone opted into email, they're fine getting email forever. But that's not how attention works. Lucas: No, it's not. And this is actually where a handful of brands are doing something smarter. Warby Parker, for example — they don't just ask for channel preference at checkout. They re-prompt during the post-purchase flow. After you receive your glasses, they ask: 'Do you want care tips via email or text?' That's a context-specific preference, not a blanket one. Allbirds does something similar with their loyalty program — they let you choose how you hear about new color drops versus how you hear about sales. Luna: So they're splitting preference by message type. That requires a more structured data model in your marketing automation platform. Lucas: It does. And most platforms — HubSpot, Marketo, Klaviyo — they support custom preference fields, but the default setup encourages a single 'communication preference' picklist. So you have to actively build out that multi-dimensional model. And then you have to make sure your workflows actually reference the right field at the right step. Luna: Which is where a lot of the leaks happen. Someone builds a workflow for abandoned cart — they set the channel to email by default, because that's the easiest thing. They don't write a conditional branch that checks: did this person opt into SMS? If so, send the abandoned cart reminder via text. Lucas: Exactly. And if they did opt into SMS, but the workflow sends an email anyway, you've just created a negative experience. They might even mark you as spam because they didn't expect to hear from you on email. I've seen cases where unsubscribe rates are actually driven by channel mismatch — people aren't leaving because they dislike the brand, they're leaving because the brand is talking to them in a way they didn't ask for. Luna: So what's the concrete fix? If someone's listening and managing workflows today, what should they look at first? Lucas: I'd say do a preference audit as part of your next quarterly workflow review. Export your active contacts and compare their stated preference field against the channel actually used in the last three automated messages they received. I bet you'll find a mismatch rate of at least 5 to 10 percent. Once you see the numbers, it's much easier to justify building conditional logic into your core journeys. Luna: And that audit itself — that's something a marketer can do without buying new tools. It's just a data export and a few hours of analysis. Lucas: Right. And the return on that time is potentially huge. If you fix channel mismatch for even 5 percent of your audience, you're looking at better deliverability, lower unsubscribe rates, and likely higher conversion. Because you're finally meeting people where they actually want to be met. Luna: It's interesting — this whole conversation makes me think about how we handle preferences here. We're listener-supported, no ads, and we've got a small group who chips in monthly through buy me a coffee dot com slash fexingo. And that model works because people choose to engage on their terms — email, podcast app, whatever. It's the same principle: don't assume, let them choose. Lucas: That's a good parallel. And it's true — a handful of listeners supporting us on buy me a coffee dot com slash fexingo is what keeps this show ad-free and allows us to dig into these topics without any sponsor pressure. So thank you to those folks. But back to the preference problem — I think there's another layer that's even more subtle. Luna: What's that? Lucas: Implicit preferences. Not what someone clicked in a checkbox two years ago, but what their behavior actually tells you today. If a customer hasn't opened an email in six months but clicks every push notification, their implicit preference is push. But most automation platforms don't have a feedback loop that updates the preference field based on engagement. Luna: So you could have someone who technically opted into email, but their behavior says they don't want it. And the workflow keeps sending email, driving them further away. Lucas: Exactly. I've seen brands build what they call 'channel scoring' — a simple model that tracks engagement by channel over a rolling 90-day window. If email engagement drops below a threshold, the system flags that contact and either shifts them to their most engaged channel or sends a re-permission prompt. It's not complicated, but it's rare. Luna: And that's something platforms like Marketo can do with a few custom fields and a scheduled campaign. You don't need AI for this. Lucas: No, you don't. And that's the thing — a lot of the fixes here are straightforward data hygiene and workflow design. They're not expensive, they're not technically hard. They just require a shift in thinking from 'I'm going to send this message' to 'I'm going to send this message to this person via their preferred channel, and I'm going to verify that preference is current.' Luna: It's a respect question. You're respecting the customer's choice of how they want to be reached. Lucas: And that respect pays off. There's a study from MarketingSherpa that found that 72 percent of consumers say they're more likely to engage with a brand that personalizes based on their preferred channel. So it's not just about avoiding mistakes — it's about actively choosing the channel that earns attention. Luna: I want to go back to the example of the brand that lost 14 percent open rates after an ESP migration. What did they actually do to fix it? Lucas: They ran a one-time campaign asking customers to reconfirm their channel preferences — not just opt-out, but actively re-choose. They sent it via all three channels: email, SMS, and push. About 18 percent of their base responded, and among those, nearly 30 percent changed their preference from what was on file. Then they rebuilt their core workflows to include a conditional branch at the first step: check preference, then route. Within two months, open rates recovered and actually improved by 6 percent over the pre-migration baseline. Luna: So the fix was a combination of a re-permission campaign and better workflow logic. And it only took two months to see results. Lucas: Right. And the cost was basically the time it took to build the re-permission email and update the workflows. No new platform, no consultants. Just a marketer paying attention to the data they already had. Luna: That's the kind of story I wish more marketers heard. Instead of 'buy this AI tool to optimize your sends,' it's 'look at your existing preferences and fix the gaps.' Lucas: Exactly. And I think the next frontier here is real-time preference sync. Most platforms still update preference fields with a delay — sometimes up to 24 hours. So if someone opts out of email at 10 AM, they might still get a campaign sent at 2 PM because the workflow pulled a stale field. Some of the newer platforms are moving toward event-driven preference updates — as soon as the preference changes, the workflow branch updates immediately. Luna: That would eliminate the lag problem. But until that's standard, what's the workaround? Lucas: A simple one: schedule your preference field sync to run immediately before your campaign sends. Most platforms allow you to set a refresh timing. If you know you're sending an email at 2 PM, set your preference sync to run at 1:55. It's a small tweak that can prevent a lot of frustration. Luna: And it's something you can implement this week. I think we've given listeners a pretty actionable checklist: audit your current preference data, build conditional routing, and refresh your sync timing. Lucas: Plus get serious about implicit preferences. That's the one that most people overlook but that probably has the biggest long-term impact. Because preferences aren't just what someone told you once — they're what their behavior tells you every day. Luna: Alright, that's a solid place to wrap. Next time, let's talk about the opposite problem — what happens when you have too many channels and no clear way to prioritize. Lucas: That's a good one. I'll dig up some examples. Thanks for listening.