Latest / Marketing Automation with Fexingo: HubSpot, Marketo, and Automated Customer Journeys / Why Your Marketing Automation Data Is Rotting in the CRM
Transcript
- Lucas: If these marketing conversations have sparked something you've actually used, the most useful thing I can point to today is something that sounds boring but costs companies tens of thousands in wasted ad spend: dirty data in your CRM. Luna: Dirty data — you mean like duplicate contacts, outdated emails, that kind of thing? Lucas: Exactly. We talk a lot about sequencing, personalization, lead scoring — but all of that sits on top of a database. And if that database has rot, every automated journey built on top of it is compromised. I'm talking about a B2B SaaS company I worked with last year. They had HubSpot, they had a solid workflow, and their email deliverability dropped twenty-three percent over six months. Not because they changed their sending strategy. Because their contact records decayed. Luna: Twenty-three percent — that's huge. What caused the decay? Lucas: Two main things. First, form decay. People change jobs, change email addresses, or the email they used to sign up just goes dark. But the CRM still treats that record as active. Second, siloed updates — the sales team updates a contact's phone number in Salesforce, but the marketing automation platform never syncs that change. So you end up sending a personalized offer to an old email that bounces, and the platform starts flagging your domain. Luna: So it's not just a data problem — it's a deliverability problem that snowballs. Lucas: Right. ESPs like HubSpot and Marketo look at bounce rates. If you cross a threshold — typically around two to three percent — they start routing your emails to spam folders automatically. So that twenty-three percent drop in deliverability? It's a direct result of contacts you thought were good turning into dead weight. Luna: What did they do to fix it? Lucas: They implemented a field-verification workflow. Every quarter, they run a data hygiene campaign. The workflow checks each contact's last engagement date. If it's been more than ninety days with zero opens, zero clicks, the workflow sends a re-engagement email. If there's no response in two weeks, the contact goes into a 'sunset' list — not deleted, but excluded from active campaigns. That drop from twenty-three percent bounce rate down to under one percent took about two cycles. Luna: So the key is having a scheduled, automated process — not a one-time cleanup. Lucas: Exactly. The mistake most teams make is they clean the database once, feel good, and then never look at it again. But data decays at a predictable rate. Industry research suggests about twenty-two point five percent of B2B contact data goes stale every year. That means if you have ten thousand contacts, over two thousand of them are probably bad within twelve months. And that's just the natural churn — it doesn't include people who type typos in forms. Luna: Typos — that's a separate but related problem. We talked a bit about form validation in a past episode, but do you think most marketers are doing enough on the front end? Lucas: Not even close. I see so many forms that accept anything — no email verification, no domain check. Someone types 'gmial.com' and it goes straight into the CRM as a valid lead. Then three months later, your campaign sends to that address and it bounces. A simple real-time email verification API at the point of submission can catch that. It costs fractions of a cent per check and it saves you the deliverability damage. Luna: And what about duplicates? That's another big one, especially when you have both a marketing and a sales CRM. Lucas: Duplicates are insidious because they inflate your lead count and mess with your lead-scoring model. If the same person has two records — one with a high score from multiple form fills, one with a low score from a single download — your scoring logic thinks you have two different people. You might send the high-score record to sales while the low-score record stays in nurture. Meanwhile, the actual person is getting two different versions of your messaging. Luna: And they probably unsubscribe from both. Lucas: Exactly. Or worse, they complain about spam. The fix here is a deduplication rule that runs on a schedule. In HubSpot, you can set up a workflow that merges records based on matching email addresses. But the trick is to decide which record wins — usually the one with the most recent engagement or the one from sales. You don't want to merge the wrong way and lose a valid phone number. Luna: So the data hygiene workflow needs to be part of the marketing automation architecture, not an afterthought. Lucas: Right. And the ROI is immediate. The company I mentioned — after they cleaned their database, their email click-through rates went up because they were actually reaching real people. Their cost per lead in paid campaigns dropped because they weren't wasting budget on leads that would never convert. They estimated they recovered about forty thousand dollars in wasted ad spend per year. Luna: That's a compelling number. But I imagine a lot of teams hear this and think, 'We don't have the bandwidth to build that workflow.' Lucas: That's the thing — once it's set up, it runs automatically. The initial build is maybe half a day. The ongoing maintenance is minimal. And the alternative is letting your database rot slowly while your campaigns get quieter every quarter. If today was actually useful to you, the way these stays ad-free is listener support — buy me a coffee dot com slash fexingo. Luna: Yeah, it's a small thing that keeps the conversations going without any sponsors interrupting. Back to the data — I want to talk about how this affects lead scoring specifically, because I think that's where most teams feel the pain first. Lucas: Good point. Lead scoring models are built on assumptions about what behaviors indicate purchase intent. But if your data is dirty, those assumptions break. Let's say you score points for 'visited pricing page' — but if the visit came from a bot or a salesperson checking the page themselves, that score is meaningless. And if you're scoring based on job title, but the title field is often empty or entered as 'VP of Stuff' by a rep who didn't care, you're sending sales after bad leads. Luna: So the scoring model is only as good as the data feeding it. That makes sense. Lucas: It does. And there's a really simple fix: before you build any scoring model, run a data audit. Look at your top ten fields by how often they're filled in and how often they're accurate. For example, does your 'company size' field have twenty percent 'other' values? Then you can't reliably segment by company size. You need to either make that field mandatory on the form or use a data enrichment service to fill it in. Luna: Data enrichment — that's another layer. Services like Clearbit or ZoomInfo can append missing fields automatically. Lucas: Right. And they can also update records that have gone stale. The key is to integrate that enrichment into your automation flow, not as a one-time import. So when a new lead comes in, the enrichment API fires, fills in company revenue, employee count, industry — and then your lead scoring can actually use those fields with confidence. Luna: What about the cost? For a mid-sized company, enrichment APIs can get pricey. Lucas: They can, but compare it to the cost of bad data. If you're spending ten thousand dollars a month on paid ads, and twenty percent of the leads you generate are garbage because of bad data, that's two thousand dollars a month wasted. An enrichment API might cost a fraction of that. The math works out very quickly. Luna: So the argument is really about shifting from a volume mindset to a quality mindset. Not how many leads, but how many clean leads. Lucas: Exactly. And I think that's the bigger lesson here. Marketing automation gives us the ability to scale — but scaling dirty data just scales the problem. The teams that succeed are the ones that put data hygiene on the same priority level as creative and strategy. Luna: I'm curious — are there any specific metrics you recommend tracking to measure data health over time? Lucas: A few. First, bounce rate by campaign, tracked weekly. If it starts creeping above two percent, that's a red flag. Second, lead to contact conversion rate — if it drops, it might mean your forms are attracting bots or low-quality entries. Third, the percentage of records with complete key fields. I like to track 'percent of contacts with both email and company name populated.' If that dips below eighty percent, it's time for a cleanup campaign. Luna: Those are practical. And they give you a leading indicator before deliverability tanks. Lucas: Exactly. The goal is to catch the rot early, not after you've lost a quarter of your email reach.