In the average dealership, 42 percent of CRM contacts are unreachable. Wrong numbers, dead emails, moved households, customers who bought elsewhere years ago. Nearly half of the database your marketing spend, follow-up process, and equity mining runs against does not exist anymore. You are marketing to ghosts.
The worst part is not the number. The worst part is that nothing in your store is built to tell you the number.
Data starts dying the day it's born
Your data begins decaying the moment it is entered. Every record in your CRM was accurate for exactly one moment: the moment someone typed it. Everything after that is erosion.
It decays in four distinct ways, and each one fails silently.
Wrong vehicle on record. The customer traded privately, sold to a national buyer, or totaled the car. Your DMS never heard about it. Your equity alerts, trade offers, and service reminders are now aimed at a vehicle that left the household two years ago.
Outdated contact info. They moved. Changed numbers. Abandoned the email address. No bounce report distinguishes "bad address" from "ignored you," so the record just quietly stops performing and nobody asks why.
Misclassified intent. "Not interested" logged when they were ready to buy. A rushed BDC rep, a bad phone connection, a misread conversation. One wrong dropdown selection and a live buyer gets buried in the do-not-work pile forever.
Silent defectors. They bought elsewhere. They service elsewhere. They are still on your active follow-up list, still receiving your offers, still counted in your "active customer" total that goes in the board deck.
Why no alarm ever sounds
A store knows within hours when a deal falls through. It knows within a day when an advisor's hours drop. It can go a decade without knowing its database is half ghosts.
The reason is structural. Every system in the store reports activity, not accuracy. The CRM reports emails sent, not emails that reached a human. The marketing platform reports impressions, not impressions on actual customers. Campaign performance drifts down a point or two a quarter, which looks like market conditions, creative fatigue, or the agency slipping. So the agency gets fired and the database stays rotten.
Bad data destroys good strategy, and it always frames someone else for the crime.
What the ghosts cost
Run the math on your own store. Take your marketing budget, your BDC payroll, and the hours your sales team spends working the CRM. Now accept that some large fraction of all of it, potentially approaching half, is spent contacting people who cannot be reached or should not be contacted.
The damage compounds beyond wasted spend. Per-contact costs are calculated against a phantom denominator, so every efficiency metric you track is wrong. Equity mining surfaces "opportunities" attached to vehicles no longer owned. And the AI tools you are being sold this year will train on this exact database. AI on bad data is just fast bad decisions, and a 42 percent ghost rate is very bad data.
Find your real number
You cannot fix what you have not measured. Three steps to a baseline.
Pull a reachability sample. Take 200 random "active" CRM contacts. Validate phones and emails against a third-party service. The hit rate on that sample is your real reachability number. Most operators guess 90 percent. Most stores test under 60.
Cross-check vehicle ownership. Run a sample of your owner base against current registration or title data. Every mismatch is a ghost vehicle generating false equity alerts.
Audit the "not interested" pile. Pull 50 records marked dead in the last six months and listen to the calls or read the threads. Count how many were actually alive. That percentage tells you how much intent misclassification is costing you.
Then fix it at the architecture level, not the campaign level. Validation against external sources on the way in. Continuous hygiene, not annual cleanups. One owned data layer where a record updated once is updated everywhere.
The ghosts are not a marketing problem. They are a foundation problem. The stores that exorcise them first are the ones whose intelligence layer actually works.
FAQ
How much dealership CRM data is bad? In the average dealership, 42 percent of CRM contacts are unreachable through the information on file.
What causes CRM data decay? Four silent failures: vehicles traded without the DMS knowing, outdated contact information, misclassified customer intent, and silent defectors who bought elsewhere but remain on active lists.
How fast does customer data decay? Decay begins the moment data is entered. Contact information, vehicle ownership, and intent all drift continuously, with no system flagging the loss.
How do I measure my dealership's data decay? Validate a random sample of active contacts against third-party data, cross-check vehicle ownership against registration data, and audit records marked "not interested" for misclassification.
Why does data decay matter for AI? AI tools train and act on the database as it exists. A database that is half ghosts produces confident, scaled, wrong decisions.
Todd Smith is the Founder and CEO of QoreAI and the author of The Intelligent Dealership: How AI and Data Transform Automotive Retail.