Spooky, Scary Skeletons in Your Marketing and Sales CRM
Most marketers worry about leads ghosting them. Earlier this year, during an audit for a B2B client, we found it was actually their CRM doing the ghosting. One contact we traced had attended a live webinar and got cut off from all marketing emails 16 days later. But sales never talked to them, and none of the automated workflows were designed to boot anyone out.
Problems like this show up in every kind of CRM. In Validity's 2026 State of CRM Data Management report, 62% of organizations said poor CRM data directly costs them revenue, and 67% said bad data has contributed to delayed or scrapped campaigns.
In this blog, we share three CRM skeletons you should watch out for, starting with the one behind that webinar contact described above.
1. A Scoring Model That Suppresses Your Best Leads
That contact that attended the webinar and got cut off shortly thereafter wasn't a one-off. Live webinars were one of the client's main sources of new business, but their engagement scoring model only counted email opens. Webinar attendance wasn’t registering in their model as proof a prospect was engaged. So, if someone attended a webinar, but didn't open the next few emails, the system marked them as unengaged and suppressed them from future sends.
New leads were also added to the "unengaged contacts" segment as soon as their record was created, before they'd even been sent anything. That segment held about 42% of the client’s database. The scoring model gave more points for receiving an email than for opening or clicking one, and score decay was turned off, so someone who engaged two years ago looked the same as someone active that week. The suppression workflow was also stuck in a loop, re-enrolling the same contacts every day because nothing told it to stop. In total, 550 contacts were enrolled in it an average of 54 times each, and one record went through a "do not email ever" workflow on three consecutive days.
The client was doing the work to get people to show up to webinars, and the workflows were hiding those same people from sales. The workflows had been built at different times, several years and a couple of CRM migrations ago. Nobody had gone back to see how the pieces all worked together. That's pretty common. When someone changes roles, or an agency relationship ends, the next person usually builds on what's already there.
To check your own, pull up your automations list (workflows in HubSpot, automations in ActiveCampaign, Engagement Studio programs and automation rules in Pardot). Sort by the “last edited” date, and look at what counts toward your engagement score.
For each automation, write down what it does, who owns it, and what would stop if it were turned off. If you can't answer those for one of them, turn it off as a test and watch lead flow and sales alerts for 30 days before you decide whether to rebuild it or get rid of it. If your score only counts email opens, add the things that actually show interest, like webinar attendance and form fills, stop giving points just for receiving an email, and turn on score decay.
2. Integrations That Are Still Running Under Someone Else's Account
Integrations get set up once and then run for years under whoever connected them. At one client, the Apollo and HubSpot integration was still connected through a past agency's account, so we reconnected it to the client's own account and resynced it. At a fintech client, a Marketo-to-Salesforce sync was updating lead status incorrectly, and tracking it down meant working with their Salesforce admin on app permissions and field mappings.
To check yours, open the list of connected apps in your CRM and see whose name is on each one. If it's someone who left the company or an agency you stopped working with long ago, reconnect it under a company account and check how its fields map.
3. Duplicate Records That Make Your Pipeline Look Wrong
In Validity's 2025 State of CRM Data Management report, companies said they lose an average of 16 sales deals per quarter because of poor-quality data. Duplicate contacts are a common cause: One person who came in through a form, a list import, and a rep's manual entry can show up as three separate records, sometimes with three owners.
For a compliance software client, we cleaned up about 1,200 company records and 2,000 contacts, using enrichment data and a deduplication tool to merge duplicates, and matched each company record to its ID in the regulator's database. After that, sales and marketing were working from the same company list.
Most CRMs have a duplicate tool built in, so I'd start there. If yours doesn't, export your contacts and look for the same email or company domain showing up more than once. And while you're cleaning, pull a list of hard bounces. Some of those people changed jobs, and you'll want to find them again; others can just be archived.
Book a CRM Audit Call With Wheels Up
If you want help figuring out where to start in your CRM, grab time with Wheels Up. If your system is in need of expert support, we’ve got you. If it's something your team can handle on its own, we'll tell you that too.
If you want to keep reading, Karla's post on tactics to retire covers replacements like multi-touch attribution and progressive profiling, and both depend on clean CRM data. Our website and martech skeletons post covers the rest of the stack.
