AI advertising in the car trade: from 20 minutes to 20 seconds per car

Your best salesperson doesn't sell in the showroom. He sells on the car portal, in the two seconds a buyer scrolls past your listing and either stops or swipes on. It's decided by a photo and a headline. And yet most dealers still spend 20 minutes per car getting it ready.

The listing is the cheapest salesperson you have

Think about the maths you already know: a lead source like a car portal costs money per click, per contact, or as a flat subscription. Whatever the model, you're paying for exposure. If the listing is poor, you're paying for exposure nobody converts on. That's money gone before the customer has even spoken to a salesperson.

Conversely, a sharp listing with clean photos and copy that hits the mark is the cheapest sales work a dealer can do. It works 24 hours a day, never gets tired, and costs no more to do well than to do poorly, if you have the right tool.

What a typical car's path to the listing looks like today

1. Photography on the lot. Often in bad weather, with reflections, shadows and clutter in the background. Some dealers rent a photo studio or light tent, others don't, because it takes time.
2. Manual photo cropping and editing. Someone needs to remove the background, fix lighting, maybe blur the plate and other vehicles in the background. Typically 5-10 minutes per car if done properly, and often skipped when things get busy.
3. Writing the listing copy. Equipment list, history, arguments for why this exact car is a good buy. Most salespeople copy-paste from the last listing and change the year and mileage. The result: generic copy that looks like everyone else's.
4. Uploading to the portals and your own website. Same photos, same text, three or four places. Manually, over and over.

Add it up, and an average dealer with 40-60 cars on the lot easily spends 15-20 hours a month just getting cars ready for sale digitally. That's a full work week spent on something that could be automated.

What AI actually does about it, and what it doesn't

AI photo cleanup takes the photos you already took on the lot and cleans them up: removes reflections and shadows, fixes light and colour, blurs or replaces the background with a neutral studio backdrop, and blurs number plates automatically. It takes seconds per photo instead of minutes, and it does it consistently, car number 40 looks just as good as car number 1, rain or shine.

At the same time, AI-generated listing copy can take the car's base data (make, model, year, mileage, equipment, condition) and write text specific to that car, not a template with the numbers swapped out. A Golf with navigation and heated seats actually gets those things mentioned first, because that's what buyers search for.

What AI doesn't do: it doesn't invent faults that aren't there, and it doesn't replace a salesperson's knowledge of the car's history or a previous customer's specific wish. It's a tool for doing the tedious work fast and consistently, not a replacement for knowing the car and the customer.

Consequence: quality stops depending on how busy you are

The biggest problem with manual photo cleanup and copywriting isn't that it's impossible to do well. It's that it gets dropped when things are busy. Friday afternoon with five new cars in, no time for a photo studio, and a salesperson who just needs the listing up before the weekend, so the photos become "good enough" and the copy "quickly written". Those are exactly the cars that sell slowest and at lower prices, because they were never shown fairly from the start.

With automated cleanup and copy, that variation disappears. Every car gets the same treatment, whether it's a quiet Tuesday or a chaotic Friday. It's not just a time saving, it's a quality guarantee that doesn't depend on how busy the lot is that day.

What should you measure?

If you want to know whether it actually works for your cars, there are four numbers worth tracking in your ad partners' stats and your own CRM:

• Click-through rate. If CTR rises when photos are cleaned up, that's a direct signal the listing is competing better in search results.
• Days on lot per car. How many days pass from the car being listed to it being sold? That's the number that directly affects capital tied up and lot space.
• Contact rate. Of those who click, how many message or call? It tells you whether the copy answers the questions the buyer would otherwise have asked.
• Internal time per car. Simple but often overlooked, how many minutes/hours do you collectively spend preparing the listing, before and after you change the process?

None of the four numbers lie. They show whether the change actually moves something, or whether it's just a new tool that feels nice to have.

How to get started

1. Spot-check 10 active listings. Are the photos fair to the car? Is the copy specific, or could it apply to any car of the same model?
2. Time the process today. Stopwatch from the car arriving to the listing going live, three times. That gives you a real baseline to compare against.
3. Test on a subset of the lot first. Run 15-20 cars through automated cleanup and copy, and compare days-on-lot and contact rate against the rest of the lot over 4-6 weeks.
4. Follow up in your CRM. Look at the features that tie listing data together with lead and sales data, so you know not just what got clicked, but what actually got sold.

Follow updates on new AI features for advertising on the news page.