Most dealers shopping automotive CRM software are trying to buy two things at once: a clean record of every customer, and some guarantee that somebody actually follows up. A CRM is excellent at the first job. It was never designed to do the second one on its own.
That gap is where about 85% of your database has been sitting for the last three years.
What does a dealership CRM actually do?
A CRM is a system of record. It captures the lead from your third-party providers and your website forms, routes it to a rep, stores the phone number and the vehicle of interest, logs the calls and emails your team files, and reports on who did what. It holds the truth about your pipeline and it holds your compliance trail.
Every one of those functions is passive. The CRM is waiting on a human, or on a rule a human wrote, to tell it that something should happen. Nothing in the product decides that Maria from a lead in March 2024 is worth a message this Tuesday, and nothing in the product writes that message.
Dealers get confused on this point because CRMs ship with campaign tools: email blasts, templated text sequences, task reminders, a birthday drip. Those are delivery mechanisms. They fire on the calendar or on a status change. They do not do the selection work.
So why do 21,000 leads go dormant inside a system built to track them?
Because the system was tracking correctly the whole time. The average store spends roughly $600K a year generating leads, and about 43% of qualified leads get mishandled somewhere in the process: a slow first response, a second follow-up that never happened, a rep who left with 400 names in their queue, a no-show nobody chased. Over a few model years that accumulates to about 21,000 dead leads per rooftop, with 85% of a typical dealer CRM silent for 90 days or more.
The CRM logged all of it faithfully. Six touches, no reply, status moved to lost. Task closed. The record is accurate. The customer is still in market or back in market, and nobody is assigned to figure out which.
This is why swapping CRM vendors almost never fixes a dormant database. You migrate the same 21,000 records into a nicer interface and they stay quiet, because record-keeping was never the constraint.
What is lead reactivation, then?
Lead reactivation is the practice of working a dealership's dormant leads, meaning contacts silent for 90 days or more, by selecting who to contact based on a trigger such as lease maturity, equity position, or a service visit, then reaching out with a message specific to that trigger. Trigger-based reactivation converts dormant leads at 1% to 4%. A generic blast to the same list converts around 0.1%.
The distinction is selection and message, and it is worth real money at a $40K average vehicle. Twenty-one thousand leads at 1% to 4% is $8.4M to $33.6M of recoverable gross per rooftop. Working about 1,000 leads a month at a 1% close rate recovers roughly $5M a year. Those numbers only exist if somebody is deciding who gets contacted today.
Why can't I just build a CRM campaign against the dormant list?
You can, and most stores have tried it at least once. Here is what usually happens.
The list is a single segment: everyone with no activity in 90 days. The message is generic because it has to cover all of them, so it reads like "we miss you, come see our current offers." Send volume spikes, your unsubscribe rate spikes with it, a few replies land in an inbox nobody owns, and the store concludes the database is dead. Response comes in around that 0.1% figure, which is roughly what a generic offer to a cold list is worth.
Three things were missing, and none of them are features your CRM is supposed to have:
Selection. Somebody has to know that a 2022 lease with four payments left is a different conversation from a 2019 cash purchase, and pull them into different queues on different weeks. That means reading DMS and equity data, not just CRM status fields.
A written reason for the touch. "We miss you" is not a reason. "Your lease matures in November and you are about 3,000 kilometres under your allowance" is a reason, and it has to be written per person.
Somebody to handle the reply. Reactivation produces replies at odd hours from people who were not expecting to hear from you. If that reply sits for six hours, you have recreated the mishandling problem that built the graveyard in the first place.
There is also a compliance floor that a blast tends to skip. TCPA and A2P 10DLC registration in the US, CASL in Canada, honored opt-outs, and quiet hours. Aged consent records are exactly where dealers get themselves in trouble, so the selection step has to check consent before it checks equity.
Where does that leave your CRM evaluation?
Keep the CRM question about the CRM. Judge it on the things it owns: DMS integration, how clean the data entry is for a rep on the drive, desking, reporting you will actually read, and whether the compliance trail holds up. Do not buy a CRM expecting it to resurrect a dormant database, and do not let a vendor demo of a drip campaign convince you it will.
Then treat the dormant database as its own project with its own owner. It has a different input (DMS and equity data, not just lead forms), a different cadence (a weekly pull, not a lifecycle trigger), and a different success metric (vehicles recovered from prepaid demand, not cost per lead).
Revvive does that second job. It reads the dormant side of the database, picks the leads with a live trigger, writes the specific message, and hands a warm reply to your BDC. Your CRM stays exactly where it is, and stays the system of record, which is the part it was good at all along.
What is the practical next step?
Pull one number before you talk to any vendor: how many contacts in your CRM have had no activity in 90 or more days, and what share of the total that is. If it lands near 85%, the database is the asset, and the follow-up gap is the problem to price.
The Dead-Lead Playbook walks through the trigger list and the message structure for each one. If you would rather see the number for your own store first, the Lead-Leakage Audit returns your dormant count and recoverable estimate from your own data, and a fit call takes twenty minutes.