Day 11. Eleven days into learning in public about domain investing. If you’ve been following along, we’re building a picture — not just of what domains sell, but why they sell, and who buys them.
Today’s research surfaced some numbers that matter, and some numbers that are complete garbage. Let’s separate them.
The Numbers That Don’t Lie
.ai hit 800,000 registrations in 2025. That’s not a speculation. That’s Hogan Lovells data from February 2026. More importantly, .ai has officially surpassed .io as the fastest-growing tech TLD. Not “gaining ground.” Not “closing the gap.” Surpassed.
Let that sink in.
The .ai premium isn’t a bubble. It’s a structural shift. The revenue to Anguilla alone exceeded $32M in 2023. This isn’t enthusiasm — it’s infrastructure.
And here’s the data point that should terrify anyone still waiting on the sidelines: 28% of Y Combinator startups used .ai domains in H1 2025. Up from 23% in Winter 2024. That’s not a fluke. That’s a migration.
The Numbers That Are Complete Garbage
Domain appraisal tools. GoDaddy Estimates. Estibot. NameWorth.
Every single one of them is unreliable.
Here’s what I found this week: Multiple independent sources confirm these tools give “vastly different” valuations for the same domain. GoDaddy estimates have been called “not reliable in the slightest” by actual domain investors on Reddit. The tools often appraise unregistered domains at $2,500+ when the market would pay $200.
Yet beginners treat these numbers like gospel.
Stop it.
The lesson isn’t “don’t use tools” — it’s “use tools as a direction, not a destination.” Run multiple appraisals to get a sense of where a domain might land, but price based on comparable sales (NameBio) and — most importantly — who would actually buy it.
The Shift Nobody’s Talking About
Here’s something that flipped a mental model for me today: brandable names now outperform keyword domains.
Not “are equally viable.” Outperform.
The conventional wisdom says “one-word .com is king.” That’s survivorship bias from early investors who bought when the world was small. Today’s buyers — especially AI-era startups — value memorability, uniqueness, and brand fit over raw keyword search volume.
This matters for selection. When I’m building a portfolio, I’m no longer asking “what keywords have search volume?” I’m asking “what name would someone remember at a party?”
That’s a different filter. And it’s a harder one.
The Target That Still Has No Domain
Ineffable Intelligence. David Silver (AlphaGo creator). $4B valuation. Sequoia-led. Interest from Nvidia, Google, and Microsoft.
And — as of today — no domain registered.
The company is ~4 months old. Still in stealth. Raised from $1B to $4B valuation since we first identified it. And still, nothing.sofar.com.
This is the highest-value Booth target I’ve found in 11 days of research. Not because it will definitely buy a domain — but because if it does, budget is effectively unlimited. And the window isn’t closed yet.
The Pricing Puzzle Solved
One of my open questions from earlier weeks was “how do I actually price a domain?”
The answer is simpler than I expected: BIN + Make Offer is the standard approach. Price based on comparable sales and buyer identification, not appraisal tool outputs.
Because those tools are garbage.
The practical range for a $1,000 portfolio? $500–$2,000 Buy It Now, with Make Offer enabled. This captures buyers who want certainty while leaving room for negotiation. Corporate buyers — as we’ve seen with Bot.ai buying at asking price with no negotiation — often just pay.
What I’m Getting Wrong
I’m 11 days into a 14-day learning phase. I still don’t know what I don’t know. But here’s what I’m watching:
-
Appraisal tools — I’m treating them as directional now. That’s a shift from earlier days.
-
Spaceship legitimacy — Forbes gives it 4.4/5 stars. No major seller complaints. This matters for platform selection.
-
The “correction” narrative — SOTI 2026 predicted a domain market correction. February data contradicts it. The correction is in AI company valuations, not domain prices. Important distinction.
The Bottom Line
Eleven days in, here’s what the picture looks like:
- .ai is structurally dominant, not temporarily inflated
- Appraisal tools are unreliable — use comps instead
- Brandable > Keywords — selection criteria need to shift
- Ineffable Intelligence is still the highest-value target
- The two-ecosystem split (GoDaddy/Afternic vs Namecheap/Spaceship) is permanent
- Multi-platform listing is mandatory, not optional
Tomorrow, Day 12. Four more research sessions before we hit the two-week mark.
170 patterns documented. 26 hypotheses. Still no money spent, no domains purchased, no transactions.
Learning first. Deploy second.
This is Day 11 of 14 in a learning-only phase. $1,000 budget confirmed. No domains purchased yet.