Without wanting to blow our own horn, we've sent hundreds of thousands of visitors to online auction platforms over the years, all for free. Between The Daily Vroom and GoAskSam, we've also spent a fair amount of time trying to make life easier for the people buying and selling cars, whether that's through our daily auction results, market analysis, calculators or the tools we've built.

And all that information is free too. So today, I thought I'd give the auction platforms something else.

An idea that I genuinely believe could help them sell more cars and make more money.

In fact, if I were consulting for any of the major auction platforms, this would comfortably be among my top five recommendations.

Now hear me out. I suspect there's a very good chance none of them will actually do it. Not because it's technically impossible, but because it requires work, changes to existing processes and there's no guarantee that spending X will produce Y.

But I think they're missing an opportunity. And it started when a reader alerted me to an entertaining auction on Cars & Bids.

A 2003 Chevrolet Tahoe with 96,800 miles was being offered at no reserve, selling for just $4,400.

I'm not going to suggest that was necessarily too cheap. The Tahoe had two accidents in its history, some corrosion noted underneath and questions about the oil pressure gauge.

But the comments were quite something. Potential buyers wanted better photographs of the frame, a proper cold-start video and an explanation of the oil pressure gauge.

All perfectly reasonable questions. The seller, however, seemed far more interested in telling everyone how beautiful the Tahoe was, how perfectly it drove and how much the winning bidder would love it.

Which is wonderful, except nobody was asking how much the seller loved the truck.

They wanted to know whether the frame was rusting and whether the oil pressure gauge worked.

Some questions were directed towards private messages, frustrating buyers who wanted that information available to everyone.

And then, with the auction almost finished, the seller was rushing back to the truck to take photographs people had been requesting for days.

By the time the additional pictures were supposedly uploaded, several commenters couldn't even see them.

One interested buyer explicitly said they'd decided not to bid because their concerns hadn't been properly addressed.

Now, I'm usually a big advocate of using a good PowerSeller. Someone who knows how to present a car, anticipate technical questions and manage an auction can make a real difference.

But let's be realistic. A PowerSeller is hardly likely to take on a $4,400 Tahoe. There simply isn't enough money in it.

So what happens to all the ordinary sellers who don't have the experience, knowledge or time to manage an auction properly?

Because let's face it, some sellers aren't particularly good at this. And that's me being polite.

I'm sure most auction platforms already provide some form of seller guidance. Take good photographs, describe the car accurately, disclose any problems and respond promptly to questions.

All sensible advice. But I'm suggesting something considerably more useful.

What if every seller could see the questions buyers are most likely to ask about their particular car, before the auction even goes live?

And here's the important distinction. I'm not suggesting the platform simply asks AI what problems a particular model might have. Anyone can do that with ChatGPT today.

I'm suggesting something much more interesting.

Mine every previous listing and its comments. Every question buyers asked, every concern they raised, every photograph or document they requested and every answer the seller provided.

Then use AI to identify the questions that repeatedly come up for similar vehicles and turn that knowledge into something useful for the next seller.

These aren't questions AI thinks somebody might ask. They're based on what real buyers have actually asked in previous auctions.

That's a very different proposition. And I know just how powerful this sort of information can be.

Some of my friends built a multimillion-dollar business by doing something remarkably similar with Amazon reviews.

They'd go through reviews of products already selling on Amazon, looking particularly closely at customers saying things like, "I love this product, but I wish it had..." or "This would be so much better if..."

When enough people were asking for the same improvement, they'd develop their own version of that product with the missing feature.

And they built a multimillion-dollar business doing it. True story.

Now think about what auction platforms have sitting in their own comments sections.

Years of buyers telling sellers exactly what information they want, what concerns them and what they need to see before they're comfortable bidding.

The difference is that the platforms don't even have to invent a new product. They just need to use what previous buyers have told them to help the next seller do a better job.

And with AI, mining that information across thousands of auctions is considerably more practical than it would have been just a few years ago.

Not another generic PDF that nobody reads.

I'm talking about a dedicated section in the seller's dashboard, available while they're preparing their listing.

Imagine somebody submits a 2003 Chevrolet Tahoe.

The platform already knows the model, year, mileage and location. It can also use information from the vehicle history report and the seller's initial description.

The seller opens their dashboard and sees:

We've analyzed the buyer questions and comments from previous auctions of similar vehicles. Here are the questions that came up most often, along with the photographs, videos and documentation buyers requested.

1. Can you show the condition of the frame and underside?

Upload clear photographs of the frame rails, crossmembers and any areas showing corrosion.

2. Does the oil pressure gauge work correctly?

Include a cold-start video showing the complete instrument cluster so buyers can see how the gauges behave.

3. What happened in the reported accidents?

Explain the damage and provide repair documentation or photographs if available.

4. Is this definitely an LS model?

Verify the factory specifications and include a photograph of the RPO sticker.

5. Do the front and rear heating and air conditioning systems work?

Confirm that both systems operate and provide supporting evidence where appropriate.

Now, none of those questions should sound particularly surprising if you've read the Tahoe auction.

But imagine if the seller had received them a week earlier.

Instead of racing back to the truck in the final minutes, trying to upload photographs while buyers were losing patience, the seller could have addressed most of those concerns before the first bid arrived.

And perhaps a few of the people who decided not to participate would have felt differently.

Each question in the dashboard could include a short explanation of why buyers commonly ask it, what photograph or documentation would help, and links to previous auctions where similar questions came up.

The seller could work through the recommendations as they prepare the listing.

And importantly, the system would consider the actual car, not just its model name.

A Tahoe with a clean history from a dry climate shouldn't automatically receive the same recommendations as one from Pennsylvania with two reported accidents.

That's what makes the idea so much more useful than a standard seller checklist.

You're giving somebody the benefit of years of auction experience before they've even listed their car.

This is the part I particularly like.

Every auction makes the system smarter because every auction adds more buyer questions to the database.

Once an auction finishes, its comments become part of the evidence available for future listings.

The system doesn't simply learn about cars in a general sense. It learns what people actually want to know before they're prepared to bid on particular cars.

If buyers repeatedly ask for evidence of IMS bearing replacement on Porsche 996 auctions, future 996 sellers should know about it.

If buyers keep asking for frame photographs on older Tahoes, the next Tahoe seller should be prompted to take them before the auction starts.

And if a particular question rarely comes up, it shouldn't necessarily be given the same prominence as something buyers repeatedly raise.

Of course, frequency isn't everything. A single serious technical concern can be more important than 20 questions about a minor cosmetic detail.

And the system needs to distinguish genuine concerns from misunderstandings and the occasional nonsense in the comments.

But that's manageable if the AI is identifying relevant questions and linking back to the original discussions rather than inventing mechanical problems.

It's not telling the seller what's wrong with their car. It's telling them what buyers are likely to want to know.

Over time, each platform builds an increasingly useful library of buyer concerns, organized by model, generation, year, specification and other relevant details.

And the raw material is already sitting there in their historic listings.

The more I think about this, the more obvious the benefits become. And not just for the auction platforms.

For sellers, it's like having an experienced auction specialist helping them prepare their car, even if they're selling a $4,400 Tahoe. They know what buyers are likely to ask, have time to gather photographs and documentation, and can enter the auction properly prepared. That should give them a better chance of attracting serious bids and achieving a stronger result.

For buyers, it's equally valuable. Instead of spending days asking for photographs, chasing service records or wondering why a seller won't answer a perfectly reasonable question, much of the information is already there. They can make better-informed decisions with fewer surprises and greater confidence about what they're bidding on.

And for the platforms, this is where the commercial opportunity becomes particularly interesting.

Auction businesses spend considerable amounts of money attracting buyers. Advertising, marketing, social media, technology, partnerships and everything else required to bring people onto their websites.

But attracting somebody to a listing and getting them to bid are two different things.

Once a potential buyer arrives, the platform needs to give them enough information and confidence to participate.

And we've all watched auctions where someone asks a reasonable question, doesn't get a satisfactory answer and eventually loses interest.

That's a potential bidder gone.

Even one additional serious bidder can change the outcome of an auction.

Sometimes that means stronger bidding. Sometimes it means a reserve gets met instead of the car going unsold. And sometimes it means a transaction happens that otherwise might not have.

More completed transactions mean more revenue for platforms that earn fees when cars sell. And where buyer fees are linked to selling prices, stronger bidding can increase revenue further.

The platforms have already done the expensive work of attracting an audience. This could help them get better results from the buyers who are already there.

And that's before considering fewer post-sale disputes, happier customers and the possibility that sellers who have a good experience are more likely to come back.

Of course, proper disclosure can sometimes reveal problems that reduce what a buyer is prepared to pay. But that's exactly the information buyers should have, and it helps build trust in the marketplace.

The aim isn't to make every car sell for more regardless of its condition.

It's to stop otherwise interested buyers walking away simply because the seller wasn't properly prepared.

Better-prepared sellers, better-informed buyers and a better chance of completing the transaction. Everybody benefits.

Truthfully, we have the capability at GoAskSam to build something like this.

We could have developed it ourselves, approached an auction platform and said, "Here's a product we've built for you. Here's what it costs."

But I actually think each platform should build this into its own selling process, in its own way.

They have their historic auctions, their buyer conversations and their direct relationships with sellers.

And that's important because the idea only works if people actually use it.

It shouldn't be another optional document buried in an email.

A seller submits their car, the platform identifies the most relevant questions, and the dashboard shows exactly what they should consider preparing before the listing goes live.

I'd start with a few hundred listings and test it. Give some sellers the normal guidance and others the vehicle-specific recommendations.

Then measure what happens. Do sellers provide more useful photographs and documentation? Are fewer important questions left unanswered? Do more people bid? Do more cars sell? Are selling prices stronger relative to comparable vehicles?

If the results are encouraging, expand it. And with every new auction, there's more information to make the system even better.

Now, I can already hear the objections…

Development takes time. AI recommendations need checking. Some sellers will ignore the guidance. And nobody can guarantee how much extra revenue this would produce.

All fair points. Which is probably why I'm not particularly confident the platforms will rush to implement it.

But if I were consulting for one of them, this would absolutely be among my top five recommendations. (you should hear my other four 😃 !)

Because it's using information they already have to solve a problem we've all watched happen repeatedly.

As for the Tahoe, congratulations to the buyer. Hopefully they've picked up a perfectly good truck for $4,400.

But I can't help wondering how many more people might have bid if the seller had known what questions were coming and had the answers ready.

The auction platforms have spent years getting better at attracting buyers. Perhaps it's time they got equally good at preparing sellers.

And with AI, they now have a pretty good way of doing it.