Latest / Marketing Automation with Fexingo: HubSpot, Marketo, and Automated Customer Journeys / How Marketing Automation Personalization Costs Are Hidden
Transcript
- Lucas: You know that moment when you're building a welcome email in HubSpot or Marketo and you drop in a personalization token — first name, company name, maybe a dynamic product recommendation — and it just works? One click, one curly brace, and you feel like a genius? Luna: Oh, absolutely. That little dopamine hit of seeing 'Hi, Luna' render perfectly in the preview pane. Lucas: Right. But here's the thing — that token is actually a liability. And every marketing team I talk to underestimates the hidden cost of maintaining those personalization elements across their entire automation ecosystem. Luna: Hidden cost meaning engineering time? Or data quality issues? Lucas: Both. But I want to drill into a specific number today. A consultancy called Etumos did an audit of a mid-market B2B company running both HubSpot and Marketo. They found that the company had 1,247 unique personalization tokens deployed across 890 automated emails, SMS messages, and web forms. 40 percent of those tokens referenced fields that were no longer populated in the CRM. And 25 percent of the automation workflows using those tokens had at least one logic branch that would never fire because the underlying data was stale or misconfigured. Luna: So nearly two thirds of their personalization effort was either dead weight or actively broken? Lucas: Exactly. And here's the real gut punch — when they calculated the engineering cost to build and maintain those 1,247 tokens, including the time marketing ops spent troubleshooting broken renders, the annualized cost was about seventy-eight thousand dollars. For one mid-market company. Luna: Seventy-eight thousand? That's more than a junior marketing ops salary, just to keep personalization from breaking. Lucas: Yeah. And the kicker is that most teams don't track this. They see the shiny dashboard showing that personalized emails get 26 percent higher open rates — which is true, by the way, according to a 2025 Litmus study — but they never look at the cost side of that equation. Luna: So what causes this debt to pile up? Is it just teams adding tokens without cleaning up old ones? Lucas: That's part of it. But there's a deeper structural issue. Most marketing automation platforms treat personalization as a template-layer concern. You create a reusable email module, drop in a token like 'company name' or 'recent purchase category,' and that token is hard-coded to a specific CRM field. If the field gets renamed in Salesforce — which happens constantly during CRM migrations or cleanups — every single template using that field breaks silently. Luna: Silently meaning the email still sends, but the token renders as a blank space or a raw field name? Lucas: Or worse — it defaults to a fallback value that may be misleading. I've seen cases where a 'first name' token defaulted to 'Valued Customer' because the field was empty, and the email ended up sent to a C-suite executive who'd been a customer for six years. That erodes trust fast. Luna: So the problem is that the marketing team owns the template, but the CRM team owns the data schema. And they don't talk to each other. Lucas: Exactly. And the typical fix teams reach for is a data dictionary or a field-mapping document in a shared Google Sheet. But those documents go stale in about a month. The better solution I've seen is to introduce what I'd call a 'personalization audit SLA' — a quarterly review where marketing ops runs a script that checks every active token against the CRM schema and flags fields that have been deprecated, renamed, or are returning null values more than five percent of the time. Luna: That sounds like something a tool like Workato or Tray.io could automate. But doesn't that add another layer of complexity? Lucas: It does. But the alternative is accumulating what I call 'personalization debt' — analogous to technical debt in software engineering. Every token you add without a cleanup plan increases your maintenance burden. Over time, that debt compounds. A study from Gartner in late 2025 estimated that marketing automation personalization debt costs companies between 4 and 7 percent of their total MarTech spend annually. For a company spending two million on MarTech, that's eighty to one hundred forty thousand dollars in invisible waste. Luna: And that's before you factor in the opportunity cost of a broken personalization — lost revenue from a bad customer experience. Lucas: Yes. And I think that's the part that's hardest to measure but most damaging. If a key account executive receives an email that says 'Hi, ' because the token broke, that's not just a technical glitch — it's a relationship failure. And in B2B, where relationships are everything, that's costly. Luna: So what can a marketing ops person do this week to start cleaning this up? Give me three concrete questions to ask. Lucas: First question: What percentage of your personalization tokens are referencing fields that have a data population rate below 95 percent? If a field is empty more than five percent of the time, that token is a liability. Luna: And you can pull that from your CRM's field usage report. Lucas: Exactly. Second question: When was the last time you ran a full audit of all active automated emails and checked each token's render rate? Not send rate — render rate. How many of those emails actually displayed the personalized content correctly on the recipient's device? Luna: And I'm guessing most platforms don't surface that natively. Lucas: They don't. You have to pull email interaction logs and join them with token-rendering data. It's a SQL query, not a dashboard. But it's absolutely worth doing. Luna: And the third question? Lucas: Third question: How many of your personalization tokens are used in more than ten distinct workflows? Because if a token fails, it doesn't break one email — it breaks ten. That's your blast radius. And in the Etumos audit I mentioned, the average token was used in 17 workflows. So a single broken field could corrupt 17 different customer touchpoints. Luna: That's scary. But it's also actionable. And I think that kind of specificity is exactly what our listeners need. Lucas: If this episode has sparked something you've actually used — maybe you're already thinking about which tokens in your own system might be broken — that's exactly the kind of practical takeaway we hope for. And the reason we can keep doing these deep dives without ads is listener support. So if today was useful, you can buy us a coffee at buy me a coffee dot com slash fexingo. Luna: Yeah, it genuinely helps us keep the show focused on the stuff that matters, not on selling you anything. Lucas: Okay, back to the audit. Once you've identified your high-risk tokens, the next step is to decide whether to fix them or kill them. And I think the instinct is always to fix — but sometimes the smartest move is to deprecate a token entirely and replace it with a simpler, more reliable personalization, like a general greeting based on segment rather than name. Luna: So trade specificity for reliability. Because a generic greeting that works is better than a personalized one that's broken half the time. Lucas: Exactly. And that's a hard conversation to have with stakeholders who love seeing their name in an email. But you can show them the data: broken personalization reduces click-through rates by an average of 12 percent, according to a 2025 study from Mailjet. So a working generic greeting actually outperforms a broken personalized one. Luna: That's a powerful argument. I'm going to use that next time a VP of Sales asks why we removed the company name from the automated nurture sequence. Lucas: And that's really the core of this episode: personalization is not free. Every token has a carrying cost. And the most mature marketing teams are the ones who manage that cost proactively — through audits, SLAs, and a willingness to simplify. Luna: It's like a garden. You can plant a lot of flowers, but if you don't weed, you end up with a mess. Lucas: Exactly. And the weeds grow fast.