GoAskSam: what it is, what it isn't, and what's coming

For anyone new here, GoAskSam is our free tool that tells you where to sell your car and why, based on hard factual data, not a guess. And because a few of you have been asking the same question and I want to answer it properly.

How it actually works.

Every recommendation you get, every data point behind it, is NOT AI.

It comes from an in-house algorithm we built. Not a model guessing. There's a sales database behind it, every completed sale across every platform and auction house we track, and a rules engine on top: it resolves your exact car, pulls the comparable sales, and ranks venues on what the data actually shows, where cars like yours are selling, what they're bringing, how recently. That's the simplified version of the engine, and it's ours, and no recommendation you see ever came from an AI model.

The bits of conversation in between, understanding what you typed, asking a quick follow-up, answering a question about your result, that's AI. It's there to get you to the recommendation smoothly. It never makes the recommendation itself.

So when it tells you to sell your 993 on Cars & Bids, that came from our algorithm reading the data, not from anything generated. If the data’s thin for your car, it says so. If it can't make the call, it says that too. I'd rather tell you nothing than tell you something made up, and that's the whole product in one sentence.

The auction houses are now in.

Until recently the record covered the online platforms: Bring a Trailer, Cars & Bids, PCarMarket, Hagerty, Hemmings, Collecting Cars and the rest. As of this week it also covers the major live houses: RM Sotheby's, Gooding, Bonhams, Broad Arrow, Barrett-Jackson and Mecum.

That changes what the tool can tell you at the top end. Type in a 1964 Ferrari 250 GT Lusso and it won't point you at an online listing. It'll tell you the record shows Lussos going through RM Sotheby's, with Broad Arrow, Gooding and Bonhams also taking them, show you the actual sales behind that with the hammer price and what the buyer paid, and give you a door to a consignment enquiry.

If your car has sold before, give it the VIN.

If your car has ever crossed the block at any of the venues above, online or live, type the VIN (or the chassis number on an older car) instead of the year, make and model. If it's in the record, it'll tell you it knows that exact car, show you what it sold for and when, and go from there.

We're still in beta, and that's deliberate.

GoAskSam is open to this community and nobody else right now. That's on purpose. You're the people who'll notice when something's off, and several of you already have. If you spot an error, a bug, or a car it gets wrong, email [email protected] and tell us what you typed and what you saw. Every one of those has made it better.

What's next.

Something big. It's been in the works for years, it's the thing I actually started all this to build, and it's close. Can you guess? More very soon.

Try GoAskSam.com

Two weeks ago, Doug DeMuro put his 1999 Mercedes-Benz G500 Cabriolet up for sale on Cars & Bids and the bidding went absolutely nuts almost immediately.

Day one from the finish it was already at $125,000. By Day 2 it had reached $200,000. It eventually climbed to $209,000 and then, for what felt like forever, just sat there. When the auction finally ended, after more than 110,000 views, 24 bids and all the attention you would expect around Doug selling one of the cars he was probably most associated with, somebody came in with one final bid.

$210,000. Sold.

A perfectly respectable result, but after everything that had happened at the beginning of the auction, the ending was a bit of a damp squib. Even the comments immediately after the sale were full of people calling it anticlimactic, boring and wondering where the bidding war had gone. One commenter made an interesting point afterwards: perhaps getting to such a big number so early had actually hurt the auction. The G500 had effectively shown its hand in the first couple of days, and for the rest of the week everyone was staring at a number that was already very close to where the car ultimately sold.

Now, less than two weeks later, Doug is selling another one of his cars. And something very familiar is happening.

His 1996 Porsche 911 Turbo, pictured above and now up for auction hit $220,000 almost immediately after the auction went live. It is now sitting at $267,000 with four days still to run, despite having received only nine bids.

So I have one question. Are we about to watch the same auction twice?

The Porsche itself is obviously a very different proposition. This is a 993 Turbo, six-speed manual, finished in Arena Red Metallic with 62,800 miles. It has spent its life in California, comes with documentation reportedly going all the way back to new, and Doug has just spent more than $24,000 sorting it mechanically. He bought it roughly a year ago, put around 3,200 miles on it, used it with his family and has now put it up with no reserve.

There is also the unavoidable Doug factor. Every time he sells one of his own cars there is a debate about whether being owned by Doug adds value. I normally think people overstate the effect of celebrity ownership, but this is slightly different. Doug has not simply owned the car quietly and appeared on the title. Thousands of people have watched him drive it, talk about it, service it and use it with his kids. The car has effectively had a very public final chapter of its history, and for the right buyer that probably does mean something.

It also gives bidders a level of transparency you rarely get. Doug has openly discussed what he paid, what he has spent and what has been done to the car. In the comments he has already pointed out that, once you include sales tax and the $24,000 service, his total investment is higher than the current bid. He says he does not expect to make much, if anything, on the car.

But the part I find most interesting has less to do with Doug or even the Porsche.

It is what happens when a major auction gets bid up very aggressively at the beginning.

Most online auctions are conditioned around the opposite. Cars sit at numbers everyone knows are meaningless for days, then the serious money appears near the end. That creates momentum, excitement and, most importantly, competing bidders who suddenly have to make decisions under pressure.

With the G500, most of that price discovery happened very early. It had already reached $200,000 by the second day, then $209,000 well before the finish. By the time the clock actually mattered, there was very little left to discover. The winning bidder added just $1,000.

The 993 may be doing exactly the same thing. But I do wonder whether there is a psychological effect.

If you are watching an auction at $175,000 and believe the car is worth $280,000, you might start participating. Once you have placed a couple of bids, watched someone outbid you and mentally started thinking of the car as yours, getting to $280,000 can feel very different.

If the auction is already sitting at $267,000 when you arrive, you are basically being asked to make the full decision immediately.

There is no warm-up. That does not necessarily mean a lower final number, but it might explain why some auctions with huge early bidding end rather quietly. The excitement happened days earlier.

So now we wait….

Doug's last auction raced to a big number early, then barely moved when the clock finally mattered.

This one has raced even higher, even faster. The real question is whether Thursday gives us the bidding war everyone expected from the G500, or whether the most exciting part of this auction has already happened.

Last week I wrote about FinalLap.com, the new auction platform that is expected to launch soon with a format that feels quite different from the online auctions we have become used to.

Rather than every car simply having its own ending time spread throughout the week, the idea is much closer to a live auction. Cars build toward a Friday event and then effectively cross the block one after another. My immediate question was whether people would actually tune in for that, because there is a big difference between following one car you might want to buy and sitting down to watch an entire online auction.

I reached out to FinalLap to find out when they will be launching, but have not heard back yet, when I do, I’ll let you know.

But while I wait for them, I did hear from plenty of you, and I think the response is probably more useful anyway.

I asked whether you would actually tune in to watch a FinalLap auction even if you weren't buying anything. The answer was surprisingly positive.

Nearly half of respondents said yes, they would tune in even if they weren't buying. Another sizeable group said maybe, depending on the lineup. Only a relatively small number said they preferred the current model of auctions ending individually.

That matters because the biggest question I had about FinalLap wasn't whether somebody could build another website to sell cars. Plenty of people can do that. It was whether you could make an online car auction feel like an event again.

Judging by your responses, there might be something here.

One reader immediately compared it to the early days of Barrett-Jackson, remembering when it was still held at the Civic Center before moving to WestWorld. Another said the concept reminded them of watching Barrett-Jackson or Mecum live, except compressed into one day rather than several. Their point was simple: even if they had no intention of buying, they would watch just to see whether the whole thing actually worked.

I thought that was interesting because curiosity is probably FinalLap's first customer acquisition channel.

People don't necessarily need to be looking for a car. They need to wonder what is going to happen next.

Several readers mentioned exactly that. The live ending, the cars moving through in sequence and the possibility of something unexpected happening were all reasons they would watch. One person who had spent years attending Manheim auctions made another good point: if FinalLap gets the pacing right, it could actually be more entertaining than some traditional live auctions, where one car can drag on forever.

But there was also a pretty consistent warning buried inside all the enthusiasm.

The cars have to be good.

One reader put it particularly well. If you're going to turn an auction into an event, there needs to be a reason to watch beyond a bunch of dealer inventory ending one after another.

No-reserve cars. Fresh trades. An interesting collection. Cars people haven't already seen advertised in five other places. Something that makes you think you might miss something if you don't tune in.

I agree with that completely. Another reader raised the reserve issue. A live format sounds exciting when cars are actually selling. It becomes much less exciting if we sit there watching car after car fail to meet reserve.

That may be the single biggest thing FinalLap has to get right.

You can create all the countdown clocks, livestreams and production you want, but if the audience quickly learns that the cars aren't really going to sell, the novelty disappears. The best live auctions work because there is genuine jeopardy. Someone is buying the car. Someone else is losing it. The number keeps moving.

There were also some practical suggestions that made a lot of sense. One reader said FinalLap should notify you when a car you're following is, say, five cars away from going live. That way you don't need to sit through the entire auction just because you're interested in Lot 37.

I really like that. The goal shouldn't be to force everyone to watch for three hours. Give me a reason to dip in for the car I care about and you might find I stay for the next five.

Of course, not everybody was convinced. The strongest criticism was one I've written about many times: audience comes before auctions.

Bring a Trailer didn't become Bring a Trailer simply because it built a better bidding page. It spent years building an audience of people who liked looking at, talking about and learning about interesting cars. Eventually that audience became an enormous marketplace.

A new auction platform doesn't inherit that.

One reader's concern was essentially that there are already plenty of places to sell cars and not enough buyers to make all of them work. Without an organic audience, why should someone consign their good car there? And without good cars, why should buyers show up?

That chicken-and-egg problem is very real. But this is also why I find FinalLap more interesting than another platform launching with essentially the same format as everyone else.

At least they're trying something different. The reaction from you was far more positive than I expected. Plenty of you seem willing to tune in simply for the entertainment, and several responses compared the concept to the feeling of watching the big live auctions rather than another online marketplace.

That doesn't mean it will work. Now FinalLap has to deliver the difficult part: the inventory, the sellers, the buyers and enough genuine action that Friday starts to feel like somewhere car people want to be.

If they can do that, they may have something.