Today's newsletter is going to be a little different.

We've been working on something behind the scenes for a long time, and this morning I finally get to show you what we've built.

If you've been wondering what those slightly cryptic messages in the TDV have been about for the past week, you can stop wondering.

While we've only been teasing it for a week, we've actually been building it for the best part of a year.

And it all started with a question...

For the last three years, I've had some version of the same one land in my inbox again and again.

I've got this car. Where should I sell it?

Sometimes the answer felt obvious. Sometimes it really wasn't. And the more auctions we tracked, the more I realized the answer shouldn't be based on which platform I happen to like, where the last similar car sold, or who shouts the loudest.

We had the data. So we decided to build something that could actually use it.

Today, we're launching the Beta of GoAskSam.com - Here’s how it works.

The idea is simple. Tell Sam what car you're thinking about selling, answer a few questions about the car and what's important to you, and the system goes to work on the data we've accumulated from tracking this market every day.

And I want to be clear about one thing.

The recommendation is never made by AI.

We've spent the last year building our own algorithm, rules and decision-making system in-house. That's what decides what data is relevant, how cars should be compared, when there's enough evidence to make a claim, when the comparison needs to be widened, how your preferences affect the answer and, ultimately, what Sam recommends.

AI has one much simpler job. It powers the conversation. It understands what you're telling us and asks for what the system needs, so you're not working through a giant form. Once it has that, our engine takes over.

AI doesn't choose the platform. It doesn't choose the PowerSeller. It doesn't decide whether reserve or no reserve looks better. It doesn't invent statistics or interpret them. Our own algorithm, rules and decision-making system do all of that, based on real market data.

That matters to us because if Sam tells you something, we want to be able to explain exactly where it came from.

And the car itself is only part of the decision.

Maybe your priority is getting it sold quickly. Maybe you're prepared to wait if the data suggests another route could produce a stronger result. Maybe you only want to deal directly with an auction platform. Maybe you'd rather hand the keys to someone else and have them handle the entire process.

Those preferences matter too.

Our algorithm starts with your exact year, model and trim and looks at where comparable cars have actually been selling. It compares recent sold results, platform performance, reserve outcomes and timing patterns. If there isn't enough meaningful evidence at that level, our rules determine how far the comparison can be widened, and we'll tell you when we've done it.

Where the data supports it, we'll go further. Is there a particular day of the week when cars like yours have historically performed better? Does the data suggest reserve or no reserve? Is one platform genuinely outperforming another for this exact type of car?

And none of this stands still. As new auctions finish, the data updates automatically and feeds back into the system. The recommendation you get today might not be the recommendation you get next week. If the market changes, Sam changes with it.

We're not trying to tell you which platform sells the most cars.

We're trying to tell you which one the data suggests makes the most sense for your car, right now.

One thing I've learned from watching thousands upon thousands of auctions is that where you sell the car is only part of the equation.

Who sells it can matter enormously.

A great car can be badly photographed, badly prepared, poorly described or simply badly managed through an auction. Equally, plenty of owners are perfectly capable of doing everything themselves and shouldn't be paying someone else to do it.

So for some cars and some owners, Sam may recommend a PowerSeller.

We've been very deliberate about who we've chosen. We're launching with only a handful, all handpicked. They're experienced, they have different areas of expertise, they're spread around the country and, importantly, we've chosen people who have the capacity to properly look after the sellers we send them.

There will be more coming, but I have no interest in turning this into a giant directory of names. If Sam recommends someone, there needs to be a reason they're there.

And we're putting the PowerSellers through the data too.

Where we have enough evidence, we'll show you how that PowerSeller has actually performed with cars like yours compared with the wider market. If they're outperforming it, you'll see by how much. If they're not, we're not hiding that either.

Most importantly, they CANNOT pay to be the recommendation.

Neither can a platform.

Nobody can buy their way into Sam's answer.

Here's how seriously we take that. During testing, someone put in a 2017 Nissan Altima.

CarMax, Carvana, Facebook Marketplace - none of them are platforms we track, recommend, or have anything to do with. But for that car, it was the honest answer, so that's the answer Sam gave. That's the product working exactly as intended.

That's exactly what I want this product to do. Sometimes the best answer will be one of the platforms we track. Sometimes it'll be a PowerSeller. And sometimes the honest answer will be neither.

Use the PowerSeller. Sell it yourself. Choose another platform entirely. Completely ignore Sam.

It's your car and your money. We're just giving you considerably more information before you decide.

At launch, we're analyzing results from Bring a Trailer, Cars & Bids, Hagerty Marketplace, PCarMarket, Sotheby's Motorsport, Hemmings and MB Market.

There's plenty we haven't included yet.

UK and European platforms aren't there yet. They're coming. Traditional offline auction houses aren't there either. They're coming too.

We'll be adding more platforms, more PowerSellers, more data and more functionality as we go. But I'd rather start where we have enough data to make genuinely useful recommendations than pretend we can answer everything on day one.

And I want to emphasize day one. This is very much the beginning.

This is the GoAskSam Beta, and we're deliberately putting it in front of the TDV community first, before the wider automotive world.

We've tested it until we're blue in the face, but I know exactly what's going to happen. Someone will put in a car or combination we never anticipated and something weird will appear.

That's part of launching something new. If you wait until software is perfect, you never launch it.

So we're ready behind the scenes to fix things quickly, listen to what you tell us and build the features that actually make this more useful.

If something looks wrong, tell us. If you get a recommendation that makes no sense, definitely tell us. And if there's something you think Sam should do that it can't yet, I want to hear that too.

Every TDV reader gets three free searches every day. Anyone else who finds it gets one search a day.

There's a reason we're starting there.

Every search is doing considerably more behind the scenes than the chat makes it look like. We're pulling from our database, finding and comparing the relevant sales, crunching the numbers and running everything through the algorithm before returning the recommendation.

That takes real computing resources and costs real money.

More importantly, we're about to let a sizeable community loose on something that until now has been tested by a relatively small number of people. I don't want to overrun the system on day one.

Three searches gives everyone enough room to properly play with it while we watch how everything holds up. We expect that allowance to grow as we see how the system handles real usage, and we're working on a Pro version for people who eventually want considerably more.

The easiest way in is the button below. Click it directly from this email (below) and we'll seamlessly log you into GoAskSam using your TDV email. No password, no code, and your three searches will be waiting.

If you go directly to GoAskSam.com instead, no problem. Just sign in using the same email address you use for TDV. We'll send you a code and you'll get the same three searches each day.

So if you’re thinking about selling a car, or just wanna see what any car you’ve had or want to have, put it in and have a play, it’s free :)

Put in something you sold last year and see what Sam would have told you. Try something obscure. Change your priorities and see if the recommendation changes.

Try to break it. And when you do, tell us - [email protected]

Because you asked. That's really it.

For three years TDV has been tracking this market every day. Along the way we built the Import Calculator, the Domestic Shipping Calculator and the Jobs Board, all free, because they solved problems I kept seeing this community having.

GoAskSam is the next step.

Instead of emailing me to ask where I'd sell your car, you can now see what the data says.

You make the decision. Sam just helps you make a better-informed one.

Oh, and one more thing. (well done for reading all the way to the end)

This is only the beginning…

We've already been building what comes next. It's considerably bigger, and you'll see it in Q4.

But first, go and try GoAskSam.