No Sale

The 1974 Carrera 2.7 had 83 comments, $73,000 on the table, and 2.5 hours left when I dropped the link, and the car didn't sell and the comment got two thumbs. It's the best test we've run.

The right comment

I'd been watching the thread before I posted. @marcoducati — 895 karma, PCA member, owner of one of the 528 matching-number cars — had already written the setup:

"As a PCA member and fortunate owner of one of the 528 matching number cars I am perplexed as to why there seems to be no registry for the US cars."

That's not the pitch for Bordmappe — that's the pitch for the reply.

I posted as a direct response: I'm building a registry like that, and here's this car as an example — factory spec, service records, a living profile tied to the VIN.

Twelve minutes later, @misaak replied. He's been compiling a list of 1974 and 1975 US Carreras for twenty years, has records for 52% of the '74 coupes and 67% of the '74 Targas, has been in touch with the PCA advocate for 1970s Carreras, and invited us to compare notes publicly, right there in the thread.

Two thumbs on our comment, three on his, and the auction closed no-sale — reserve not met at $104,000.

None of the metrics make it look like a win. It was the best one.

What four tests actually look like

I've been running these drops for three weeks and the shape is clearer now.

Test #1 was post-close — auction already hammered at $68k, 14,991 views. I replied to the seller's thank-you thread, got three thumbs, and it was quiet.

Test #2 was mid-auction on the 1987 Carrera, two days before close with $69k on the board. Eight thumbs, and @moneymaz — a scene-literate guy with 50 karma — replied publicly: "super helpful." Car sold at $109,000. That was the one that made me think this was going somewhere.

Test #3 was four days out on the 1973 911T, six thumbs, no replies. The seller signed up the next day, the car sold at $106,250, and the GA4 read afterward showed 101 unique users from BaT traffic on that page alone over the following two weeks, with an average time of two minutes and forty-five seconds. Eighteen of those users went on to explore the rest of the site and three hit the sign-in page.

Test #4: two thumbs, no sale, and @misaak.

The pattern I'd been watching — more thumbs means better test — was wrong. Thumbs measure reach within the auction thread, but they don't measure whether the right person showed up.

The guy in Knoxville

One visitor from the test #3 traffic came back three times after the auction closed — six sessions total, just under two hours on the car page. The auction had already hammered, the car was gone, and he kept reading.

I don't know who he is — he didn't email, didn't comment, didn't sign up. That texture keeps showing up though, people arriving and reading carefully and leaving without saying anything. The site is doing something, I can see the shape of it in the data, but the conversion events are mostly silence.

Silence from the right people starts to feel like something.

Three minutes

After test #3, I emailed the seller — a high-volume BaT consignor with storefronts in Portland and Phoenix, for whom I'd already built timelines on two of his cars. He replied in three minutes, said the timelines add a lot of value for buyers, and said he was definitely interested, which was the reply I'd been looking for since the first drop.

I sent him the test #3 numbers two weeks later and then pitched the thing that's been sitting on my desk: a seven-day billboard run timed to one of his auctions, with Portland screens while the car is live and Phoenix screens the week after where his shop is. The creative is the car, buyers are already on BaT, and this would reach the ones who aren't.

He hasn't replied to that one. The billboard idea isn't dead — it's just waiting for the right moment, or the right car.

The inbox I wasn't watching

Earlier this year I sent a lot of LinkedIn notes and InMails about Dropsignal and most of them went nowhere. The big swing was a note to someone senior at a large sports podcast network, where I'd made the geographic argument for sports and dropped a shared connection as a PS.

Then I stopped checking LinkedIn — not a decision exactly, more like I put my head down on Bordmappe and the weeks went. The account flow, the BaT drops, the traffic, the seller emails. Two episodes of build.

He'd replied the next day.

Two weeks later I found it, panicked, wrote back on LinkedIn, and then immediately sent a follow-up email with a calendar link. He scheduled something and mentioned he has travel coming up, so he's out for a couple of weeks.

So there's a meeting. It's just TBD.

What's next

@misaak invited a comparison of notes, and the right move is to follow up when the next ROW '74 or '75 Carrera lists on BaT, when the context is live and the thread is warm.

Four tests. The one that looked like it failed found the person who's been doing this longer than the internet has existed, and the inbox I forgot to check had a meeting waiting in it. Things are getting picked up — I'm just not always the first to know.