Not all Automotive AI is created equal

How do I choose an AI vendor for my dealership?

Ask three questions. First: does it act, or does it only report? Nearly every product marketed as AI in automotive is analysis with a chat window attached — ask the rep to show you something the system did without a human asking it to.

Second: who owns the data and the ad accounts? If the vendor holds them, your switching cost is your entire history.

Third: what happens on a Tuesday when nobody logs in? A tool that depends on someone remembering to check it gets checked for three weeks and then doesn't.

What's the best way to measure marketing performance at a dealership?

Tie spend to delivered units, not to leads. Lead volume is where most dealer reporting stops, and it's the number that misleads most — a source can generate plenty of leads and almost no deliveries, and it will look healthy the entire time.

Getting there means joining your ad platforms to your CRM and your DMS, which is precisely the join most tools never make. If your reporting can't show cost per sale by source, it's measuring activity rather than performance.

Should my dealership build its own reporting or buy a platform?

Building is cheaper than most dealers expect. Connector tools start around $19 a month and pipe cleanly into BigQuery or Looker without much work, and a capable person can stand something up in a weekend.

The cost isn't the build — it's the maintenance. APIs change, connectors break, schemas drift. If you have an analyst whose job description includes maintaining it, building is a perfectly reasonable choice. If the person who would build it already has a full-time role, it will be excellent for four months and quietly stale by month eight.

Who owns my dealership's data when I use a vendor?

Read the agreement and check two things specifically: whether you can export the full history, and whose name the advertising accounts are in.

Many agencies create ad accounts under their own manager account, which means leaving costs you your campaign history and all the platform learning that came with it. Ask for export rights in writing, and for ad accounts created in the dealership's name from the start.

Can AI actually run dealership advertising, or does it just report on it?

Both exist, and the marketing language around them is nearly identical, which is why this is hard to evaluate from a website.

The distinguishing question is what the system does with what it finds. Reporting AI surfaces an insight and waits. Agentic AI takes the action, or queues it for a human to approve. Ask for a specific example of a change the system made on its own — and ask what the guardrails are, because a vendor without a clear answer to the second question hasn't thought hard enough about the first.

What should I ask a vendor before signing?

Five questions that surface most of what matters:

• Show me something the system did without anyone asking it to.
• Which of my systems does it read — and does that include the DMS?
• Whose name are the advertising accounts in?
• If I leave, can I export everything, and how long does that take?
• What's the real channel scope, and what sits behind a higher tier?