A working guide for GMs and sales managers
Most of what gets called a dealership marketing problem is not one. The leads are arriving. What happens in the hour after they arrive, and in the week after that, is usually where the money goes. This is a guide to telling those two problems apart before you spend against the wrong one.
There are two genuinely different failures, and they need opposite responses. If not enough of the right people know you exist, that is a marketing problem and more spend may be the answer. If people are already raising their hands and going cold, spending more makes the leak bigger.
Almost every store believes it has the first problem. Most of the ones we look at have the second. The tell is simple: pull thirty leads from last month at random and read what actually happened to each one. Not what the CRM says happened — what a customer would say happened if you asked them.
Do that before you read the rest of this. It takes about an hour and it decides everything else.
Where to look
Marketing owns two of these. Your floor owns two. Management owns the last one, and it is the one most often skipped.
Are the right people finding you?
What to measure
The common trap
Judging a source on lead count instead of on leads that become conversations.
What happens in the first hour?
What to measure
The common trap
Trusting the CRM's reported response time. It usually records when someone opened the record, not when the customer heard back.
What happens after touch one?
What to measure
The common trap
Where most stores actually lose deals. Touch one is usually fine. Touch four rarely happens.
Does the unit look worth driving to see?
What to measure
The common trap
Treating listing copy as a chore. It is the only salesperson working on that unit at 11pm.
Can you tell what worked?
What to measure
The common trap
Changing three things at once, then attributing the result to whichever one you liked.
Where AI actually fits
We build this software, so treat the following as an interested party being specific rather than a neutral survey. The honest split:
The one rule
Nothing reaches a customer until someone on your floor approves it.
Every tool that sends on its own eventually sends something you would not have sent. Keep the draft automatic and the send human, and you get the speed without the exposure.
Speed and consistency of response to the leads you already pay for. Most stores buy more lead volume before they have fixed what happens to the leads arriving now. Measure your real first-response time and your follow-up completion rate after touch one before you increase spend.
There is no credible universal number, and anyone quoting one without seeing your gross, your turn rate and your market is guessing. The more useful question is what share of your current spend produces attributable, contactable leads — and what happens to those leads once they arrive.
It helps most with the unglamorous parts: drafting consistent follow-up, writing listing copy grounded in real vehicle data, and preparing responses faster than a busy salesperson can. It helps least with strategy and with anything requiring judgement about a specific customer. Keep a human approving anything that reaches a customer.
No. Unsupervised outbound messaging is where dealerships get into trouble — with compliance, with tone, and with customers. The safe pattern is AI drafts, a person reviews and sends. That is the design constraint ManifestoAuto is built around: nothing reaches a customer until someone on the floor approves it.
A layer that sits on top of the CRM and DMS rather than replacing them. It handles pipeline visibility, drafted responses, listing copy and approval queues, and it governs which models and data the store's AI is allowed to touch. It is not a replacement for your system of record.
Free, no call required
One page per quarter. It forces you to write down your baseline before you change anything, name a single owner per stage, and state what you expect to move — so that in ninety days you can tell whether it worked instead of arguing about it.
Want the baseline measured rather than self-reported? That is what the Readiness Audit does.
A fixed-fee diagnostic that produces your real baseline and a 90-day roadmap.
Readiness AuditPipeline, first-touch SLA, approvals and listings in one place. Drafts never auto-send.
Sales CRMA self-paced course for salespeople who want to use AI without getting it wrong.
The course
ManifestoAuto