Before You Buy Any Domain, Answer These Four Questions

Most domains are bought on a feeling: the name sounds good, the extension is trending, someone on a forum said it is the next big thing. I have paid for that feeling before — and the names I regret are exactly the ones where I skipped the questions below.

These are the four questions I now answer before I spend a dollar. They are boring, they take twenty minutes, and they are the difference between a portfolio and a subscription list.

1. Is it actually available?

Whois lies under load, and DNS proves nothing: thousands of registered names carry no nameservers at all. The registry is the only source of truth. For any country extension you ask the registry's RDAP endpoint directly — 404 means free, a data record means taken.

I learned this the expensive way. A bulk whois tool told me 16,643 three-letter .si names were available. Checked properly against the registry, the real number was 263. Had I trusted the first number I would have built a plan on a fantasy.

2. Who is the buyer — can you name them?

Say it out loud: "This name will be bought by ___ because ___." If the sentence needs a paragraph, it is not a business decision, it is a bet.

A service keyword passes this test easily: plasterer.si is bought by a plastering company, kindergartens.si by a group of nurseries. A long compound with the extension stitched into the name fails it — which is why I walked away from names that were available and looked clever.

3. What does it cost to keep it alive?

The purchase price is not the cost. The renewal is. Twenty names at roughly fourteen dollars a year is $280 every year, whether they sell or not. A trending extension does not rescue a portfolio of names nobody wants; it just makes the renewals more expensive.

My rule: no name enters the portfolio without a price I would sell it at, and a review date. If it has not moved in ninety days, it goes back to the market instead of into next year's renewal invoice.

4. What did comparable names actually sell for?

Not what they are listed for — what they sold for. Recorded sales for the extension I was studying this week make the market readable: recursive.si at $20,000, travelagent.si at $19,988, bot.si at $20,000, and two-letter names yn.si and xh.si in the thousands of pounds. Every one of them is either short or immediately understandable — none of them needs explaining.

That pattern is the entire strategy in one line: buy what a stranger understands in one second.

The same four questions, applied to one extension

I ran this checklist across the entire .si registry earlier this month. The result took four numbers to summarise:

• 46,066 names were added in September alone — against 3,515 in August.
• All 676 two-letter names are registered. Available: zero.
• Of 208 AI/SaaS keywords — remove, reels, upscale, prompts — available: zero.
• Of 17,576 three-letter names, 266 were free in the morning and only 28 by the evening.

That is what an extension in the middle of a trend looks like from the inside. The easy layer is gone; what remains is a smaller, slower game of service keywords, short brandables and patient pricing.

What I do with the answers

Once the four questions are answered, buying becomes mechanical. I score the name for brandability, model the return with hold costs and marketplace fees, confirm availability against the registry one last time, and only then pay. I built those steps into DNWAI so I stop doing them by hand — the brandability score, the portfolio analyzer and the ROI calculator are free to use.

And if you want to see the full research rather than the summary, the whole .si availability report is published free, including 50 names that were still unregistered when I checked.