Every Classifieds Tool Is Built for the Wrong Customer

If you've ever tried to launch a classifieds site for a niche community (classic cars, vintage watches, a car club, a hobbyist forum), you've probably hit the same wall. The tools are either too expensive, too old, or built for a completely different problem.

I build classifieds software for a living. I run Yclas. So when people come to me with this problem, I took it seriously and went looking for what's actually broken. I read hundreds of forum threads, scam warnings from car clubs, Reddit posts from founders, and comparison reviews of every marketplace builder on the market.

What I found is that the complaints cluster into five distinct pains, and they're not five separate problems. They're symptoms of a whole category of software built for the wrong customer. Here's what people are actually saying, in their own words.

“The tools are either too expensive or stuck in 2009”

This is the most common complaint, and it splits cleanly down the middle.

On one side: the modern SaaS marketplace builders. Sharetribe is the one everyone evaluates first, and the one most people abandon. Its pricing starts at $139/month, and the first tier that lets you use your own domain is $259/month, then $389/month if you want API access. On top of the subscription, you get 50 to 500 free transactions a month, and then $0.19 per transaction after that, plus Stripe's processing fees. For a hobbyist community or a niche founder validating an idea, that's a non-starter.

On the other side: the self-hosted PHP scripts. Yclas, Osclass, Flynax. They're cheap (Flynax is a $95 to $195 one-time license) and they're capable. But they look their age. As one operator I talked to put it: “Osclass looked like it was designed in 2003 and nobody had updated it since.”That's not a knock I enjoy repeating, but it's representative. People don't reject these tools because they lack features. They reject them because they'd be embarrassed to put their community in front of them.

So the market bifurcates into “too expensive” and “too dated,” with nothing in the middle that's both affordable and modern. The comparison sites (Prometora, Roobykon, Appscrip) all exist because founders are desperate enough to read eight-way feature tables trying to find it.

“I just want listings and contact. Why is everything a marketplace?”

This one is more subtle, and it's the root of a lot of wasted money.

Almost every modern marketplace builder is built transactional-first. Sharetribe's entire architecture assumes money moves through the platform: Stripe Connect, payouts, booking flows, service fees. Prometora, Arcadier, Kreezalid, CS-Cart, same DNA. They're built to be Etsy or DoorDash in miniature.

But a huge number of people don't want to process payments at all. They want sellers to publish a listing and buyers to get in touch. That's it. The transaction (the inspection, the negotiation, the wire transfer, the handshake) happens off-platform, the way high-trust purchases always have. Classic cars, vintage watches, boats, project cars: nobody wants a payment processor anywhere near a $40,000 car. Escrow would actively scare serious buyers off.

Directorist, a directory-software company, states the category split as conventional wisdom: “Build a classifieds site if you want to connect buyers and sellers without handling payments, lower infrastructure complexity.” And yet the tooling market barely reflects it. If your model is listings-only, you end up paying transaction-engine prices for machinery you'll never use, or fighting a tool whose every screen assumes checkout.

The “no checkout” requirement isn't a missing feature for these customers. It's the whole point.

“My members keep getting scammed”

This is the pain that's gotten acute in just the last few years, and it's the one that finally made me pay attention.

Trust in the big platforms has collapsed in enthusiast niches. Classic-car online scams are now at record levels. The Better Business Bureau reports roughly $45 million a year in losses to fake classic-vehicle scams, with a median loss of about $12,600 and some victims out upwards of $36,000. A veteran classic-car dealer told Spectrum News in late 2025 that scam calls from defrauded buyers “used to come monthly, now come weekly.”The playbook is well-known and depressingly effective: a seller claims to be “working on an oil rig,” can't show the car, and asks the buyer to wire a deposit to a fake “eBay Vehicle Purchase Protection” escrow account that looks real enough to fool a 72-year-old at his bank.

The clubs are the trusted layer in all this. Classics World calls car club forums “one of the greatest defences against scammers.” But the clubs' own tooling undermines them. Their classifieds run on forum threads where members' phone numbers and email addresses are posted in the open, scraped by bots, and harvested by scammers. The FBI's Internet Crime Complaint Center has recorded roughly 27,000 complaints of online vehicle-sale fraud since 2014, totaling more than $54 million in losses. The Better Business Bureau confirms virtual vehicle vendor scams hit new highs in 2023, targeting classic-car buyers in particular.

The e-commerce platforms don't help here, because their answer to trust is escrow and reviews, exactly what these buyers don't want. What they actually need is structural: a way for a buyer to contact a seller without the seller's contact details being public, until the seller chooses to share them. That single architectural choice defeats most of the scrape-and-scam playbook. Almost no classifieds tool offers it as a first-class feature.

“I'm drowning in volunteer labor”

This is the pain nobody talks about publicly, because the people suffering from it are unpaid volunteers doing it out of love.

Here's how classifieds actually work inside a typical car club today. A member wants to sell a car. They email the club's classifieds editor, often a retiree who volunteered a decade ago because somebody had to. The email arrives with four oversized photos and a paragraph: “69 Mustang, 351 Cleveland, runs good, $18k OBO, call Jim at [number].” The editor resizes the photos, formats the HTML, uploads it, links it, checks it on mobile. Twenty to forty minutes per listing. Six to twelve listings a week in season. Three to five hours every single week, unpaid, for years.

One editor I learned from has been doing this for eleven years, maintaining the site in Dreamweaver, resizing photos in MS Paint, terrified of breaking the HTML she barely understands. She's burnt out, she's asked the board for help for three years, nobody volunteers, and she's threatened to quit twice. When one of her members got scammed because his phone number was public, she felt responsible and cried.

This isn't a software preference. It's a slow-motion crisis. The volunteers keeping community classifieds alive are aging out, and the tools they're given are either too complex to learn at 73 or too expensive for a club whose dues are $30 a year. When the volunteer quits, the classifieds die, and the community loses one of the few reasons members still engage.

“Nobody comes to the site. The buyers are on Facebook.”

This is the existential one, and it's the hardest to solve.

Here's the data point that crystallized it for me. A membership chair at a BMW club listed his E46 M3 on the club's classifieds channel. He got two inquiries in a month. He posted the same car on Facebook Marketplace. He got forty inquiries in a day.

That ratio (roughly fifty to one) is why “build a modern classifieds on your own domain” keeps failing. You can ship the most beautiful, searchable, mobile-friendly site in the world, and if the buyers are on Facebook, you've built a nicer graveyard. The forum isn't dying because the software is old. It's dying because there's no traffic, because the buyers left. Putting new software on top of an empty room doesn't fill the room.

And it cuts both ways. The clubs with thriving Facebook groups have the inverse problem: ten thousand members in an unofficial group they don't own, don't moderate, and can't measure. Members get scammed by fake listings. Flippers exploit the club's brand. The club gets zero data, zero credit, zero revenue. They desperately want to bring that energy “on domain,” but they're terrified that moving off Facebook kills the reach that makes it work. As one club president told me: you are asking me to compete with a dopamine engine using a URL.

These aren't five problems. They're one missing category.

Read those pains together and a pattern emerges. The tools are too expensive or too old. They're built for transactions nobody wants. They expose members to scams instead of protecting them. They drown the volunteers who run them. And they ignore that the buyers live somewhere else entirely.

Every one of these symptoms points at the same root cause: the entire classifieds tooling market is built for the transactional marketplace customer, not the listings-and-contact customer. The person who wants to run a trusted, low-friction, reach-aware classifieds for a real community has no purpose-built tool. They're forced to choose between paying Sharetribe prices for machinery they'll never use, wrestling a 2009 PHP script into something modern, or watching their community drift onto Facebook and die there.

That's the gap. And it's the reason “how to build a classifieds site” search results all feel like they were written a decade ago. The demand is real, but the supply is aimed at the wrong buyer.

What I'm building in response

I'm not going to pretend I've solved all of this. But the pain points above have reshaped what I'm building, and each one maps to a specific decision.

For “too expensive or too old”: a modern hosted platform, priced for the actual customer. A founder tier around $150–349/month, not $259 plus transaction fees. Modern stack, not a PHP script in a new skin.

For “everything's a marketplace”: no checkout, no transaction cut, confirmed as a product stance. Listings and contact. That's the architecture.

For “my members keep getting scammed”: buyer-to-seller messaging built as a private relay. Emails and phone numbers are never public until the seller chooses to reveal them. The scam playbook breaks at the architecture level.

For “drowning in volunteer labor”:a real import and migration tool, so a club or founder can move their existing listings in one pass, and an interface simple enough that the person who's been hand-editing HTML in Dreamweaver can actually run it.

For “the buyers are on Facebook”: this is the hard one, and I'm honest that I haven't fully cracked it. The shape of the answer, though, is becoming clear. Listings live on your domain as the trust layer, and the platform pushes them outward: to Google via clean SEO pages, to Facebook via shareable cards, to an owner's email list via a weekly digest. The site isn't a destination you beg people to visit. It's the system of record that distributes everywhere, with inquiries routed back through you. “List once, appear everywhere, own the trust” instead of “please leave Facebook for us.”

I'm building this with two design partners before I scale: one in classic cars, one in a totally different niche (vintage watches) to make sure it's a platform, not a one-trick vertical. The watch moderator I'm working with needs the exact same architecture as the car enthusiasts: rich custom fields, no checkout, secure contact, provenance links. Different hobby, identical pain. That's how I know the gap is real and the category is missing.

The question I'm asking everyone

Before I build more, I'm talking to community operators and niche founders about one question, because it's the pain that decides everything else:

On day one after you launch, where does your first buyer come from?

The people with a real answer (a newsletter, a subreddit they moderate, an existing audience) are the ones this can work for. The people who wave their hands and say “I'll post it in some Facebook groups” are the ones who'll build a nicer graveyard and blame the software. That question is now my filter for everything.

If you run a niche community and you've been hand-maintaining classifieds on a forum or a Facebook group, or abandoned Sharetribe at the pricing wall, or given up on WordPress, I'd like to hear how it actually works for you, and where the buyers really are. Especially if you've been scammed, or watched a volunteer burn out, or paid $259 a month for a tool that assumed you wanted to process payments you never asked to process.

That's the pain. I'm building toward it.

I build classifieds software (Yclas) and I'm building a new platform for the listings-and-contact customer the current market ignores. If any of these pains sound like yours, I'd value 20 minutes. The reach question is the first thing I'll ask.

— Oliver Servín · oliver@yclas-neo.com