Don't fall in love with your product

The product took you a weekend. You described what you wanted to an AI, refined a few prompts, and by Sunday night it ran: landing page, signups, payments. A website built with AI, out of nothing.

The name took three weeks. Shortlists. Arguments with friends about whether this one sounds like a startup and that one sounds like a law firm. Checking domain after domain, finding each one taken, starting over on Monday.

The product sat there, finished and unpublished, while you hunted for a name worthy of it.

Here's the sentence nobody says at that table: Google cannot hear how your name sounds. And every week spent on the name was a week the product wasn't in front of anyone.

Stop loving your product. Love the problem. Ship several small things at once, keep the one strangers actually use, drop the rest. And flip the order you were taught: pick the domain first, build the product second.

That's the whole argument. The rest is how to do it in a weekend.

Code is cheap now. Being found isn't.

The AI that built your site isn't yours alone. Anyone can create software without coding — a weekend, a credit card, almost no money. That's wonderful, and it changes what's worth fighting for. When everyone can build, building stops being the edge.

What you can't generate is a place people already walk past. A brand-new domain is an empty side street at 3 a.m. The best sign in the world, and nobody walks by until the street gets known. That's the months of waiting nobody mentions.

And quiet niches don't stay quiet. The idea you're nurturing is being built somewhere else this week, probably by several people at once. Not a statistic — just what happens when making software gets this cheap.

Pets and cattle

Server operators have an old saying: pets vs cattle. A pet has a name; when it gets sick, you sit up all night with it. Cattle have numbers; when one fails, you replace it and keep moving. Brutal about animals. Useful about projects.

Most first-time founders treat their product like a pet. Every decision is heavy because there is only one. Should you build one product or many? Build a small numbered batch — three to five — and let strangers pick the winner. How many side projects should you launch at once? Enough that no single one can break your heart.

The naming trap

How long should it take to choose a startup name? An evening. Not a month of evenings.

Google doesn't rank names by how they sound. There is no factor for "feels like a real startup." A plain name on a domain with an honest past beats a beautiful name nobody links to. People forgive plain names all the time, as long as the thing works. (Trademarks are a separate matter: see Trademark in your domain before you commit to a name you can't afford to lose.)

Domain first, product second

Now the flip. The usual order: get an idea, then hunt for a domain to match it. Reverse that.

Some expired domains are corners that signposts have pointed to for years. Newspapers, universities, ordinary blogs chose to link to them — a backlink, in plain words: someone else's site pointing at yours. The shop is gone, but some of those signposts are still standing, and search engines still follow them.

So look at drops whose past has real editorial links — articles and blogs that chose to link, not profiles on big platforms like GitHub or Medium, which anyone can create in five minutes. See what the domain used to be about. Then build a small, honest product on exactly that topic. You start on a street that's already known instead of waiting for yours to become known.

Two honest caveats.

A good drop is a chance at a head start, never a guarantee. Google may still treat the domain as new. (People talk about a "sandbox" that holds new sites back for months. That's a community term; Google has never confirmed such a thing exists.)

And the domain must stay the same kind of thing it was. If the corner was a bakery for ten years, open a bakery — not a casino. Buying a dropped domain to host unrelated or low-value content, just to ride its old links, is expired domain abuse, and Google's spam policies name it as spam.

Not every busy corner is a good one, either. Some expired domains sold stolen goods: spam, casino, pills in their past. There's foot traffic, but the police are watching. Casino, pharma or adult past explains what that means for you.

Check the past before you build

A short checklist for each candidate (the full version: Domain history checklist):

• What was it about? The Wayback Machine shows you the old site. Same topic as what you'd build, or close?
• Clean history? Spam, casino, pharma in the record — walk away, however strong the links look.
• Real editorial links? Articles and blogs that chose to link, not platform profiles. Look at who linked, not how many.
• Old enough to matter? The earliest record and the drop years tell you whether the history is real or a footnote.
• Still in the index? A site: search shows what Google holds, and Safe Browsing should come back clean.

One thing nobody can see from outside: manual penalties. Only Search Console shows those, and only after you own the domain. There is no public list. Anyone who claims otherwise is selling you something.

A weekend, not a quarter

Here's the whole method as a plan:

1. Shortlist five drops at registration price. No auctions, no bidding wars. Five corners, standard fee.
2. Run the checklist on each. Keep the ones with an honest, on-topic past.
3. Build one small product per domain. A real tool that does one thing, made with the same AI that built your first one. Not a pile of doorway pages — mass-producing pages for search is its own kind of spam, and it has its own page: Build products, not pages.
4. Put each one in front of people. Don't wait for Google to send visitors; go where people already are. Start with How to launch a product people can find, submit to the ones worth it in Startup directories: which are actually worth submitting to, and if Product Hunt fits, read How to launch on Product Hunt first.
5. Wait a few weeks, then open Search Console. Impressions, clicks, queries. Numbers, not feelings — this is how you validate a business cheaply.
6. Keep the one or two that move. Drop the rest without a funeral. The batch was the point.

Where caring still belongs

You get to love something: the user and the problem. Spend your care on whether the thing takes a real ache away from a real person. That's the love that pays back — and once the batch shows you which product deserves it, you'll have months of care ready for the winner.

The name, the logo, the shade of blue in the header: later. For the one that survived.

What to do now

1. Shortlist five expired domains at registration price. Check each past with the Wayback Machine, a site: search, and Safe Browsing. Our Chrome extension, DD Backlink Strainer, checks a domain's backlink past right on the expired-domain list (free plan: 250 checks a day).
2. Build one small, honest product per domain, on the topic the domain was known for. Its old name is your name; give it no more than an evening.
3. Put each in front of people this week — one launch platform, one community, one directory.
4. Mark the date. In a few weeks, open Search Console, keep what moved, and shut what didn't.