401.6 million. That's the number Verisign put in its Q2 2026 Domain Name Industry Brief — the first time the global internet crossed 400 million domain names in one quarterly snapshot. I read Andrew Allemann's recap on Domain Name Wire before I opened the PDF, and my first thought wasn't celebration. It was: how many of those names would I actually want to own?

Full disclosure: I almost shrugged at another registration milestone. Then I checked premium .com asks on my watchlist and stopped shrugging.

Not many. Maybe a few hundred across the whole planet that clear my bar for resale or development. The rest are campaign microsites, typo blocks, parked grids, and names somebody registered at 2 a.m. because the registrar offered a coupon. I've been buying domains since before Twitter was X. These milestone reports always arrive with the same breathless headline — the internet is huge now — and the same quiet truth underneath: scarcity at the top didn't move an inch.

Q2 added 9.1 million registrations quarter over quarter and 29.9 million year over year, an 8.1% bump. .com sits around 166.6 million; .com plus .net together hit 179.1 million after growing about 3.0 million in the quarter. New gTLDs reached 52.9 million, up roughly 34% YoY on a smaller base. China's .cn reported 20.9 million; Germany's .de 18.0 million. Big denominator. Tiny trophy shelf.

What pushed us past 400 million domain names in Q2 2026?

Three engines, same as always — just louder at scale. ccTLD growth from local businesses registering in-country. New gTLD promos and defensive corporate sweeps. Steady .com churn where expired junk gets replaced by fresh experiments. ICANN's industry overview tracks the same structural pattern: the namespace expands because the economy expands, not because premium inventory multiplies.

Verisign's brief breaks the stack clearly — legacy .com/.net still anchor the chart, new extensions grow faster on percentage math, country codes fill geographic lanes. China and Germany alone account for nearly 39 million names between them. None of that hands you a four-letter .com at retail.

I've started separating "world registrations" from "my market" in every quarterly review. Analysts love the round number. Portfolio math lives in the rows underneath.

Does 400 million domain names mean premium .com gets cheaper?

No. Growth at the base and pricing at the summit run on different tracks. When a funded founder needs a brand they can say once in a demo without spelling it twice, they're not shopping four hundred million strings. They're shortlisting maybe eight. That funnel doesn't widen because .cn added another million local shops.

I checked comps on NameBio the same afternoon the DNIB dropped. Short .com brandables and obvious category names hadn't softened. If anything, sellers read "record registrations" as proof that demand is alive and nudge BINs up. I don't love that behavior. I understand it. Emotion moves ask prices faster than spreadsheets.

Confession: I almost passed on a two-word industrial .com in April because I told myself the market was "too big now." It sold privately in June above my ceiling. The 401.6 million figure didn't punish that seller. My hesitation did. Lesson learned, again.

More registrations also don't fix the patience problem on average brandables. Good names still sit until the right buyer shows up. Macro growth is not a liquidity wand.

TLD classApprox. scale — Q2 2026 —Growth signal for buyers .com alone~166.6MStill dominates trust; premium tier tight .com + .net~179.1M — +3.0M Q2 —Legacy liquidity; not a discount signal New gTLDs~52.9M — +34% YoY —Surface area for experiments; selective trophies .cn / .de leaders20.9M / 18.0MLocal trust; weak comp for US brandables Global total401.6M — +8.1% YoY —Activity headline — not premium supply

Read the table once, then forget the total row. The rows that matter for your wallet are .com and the few alt-TLDs your buyer persona actually types without thinking.

Where's the scarcity if 400 million domain names exist?

At the pronunciation layer. Four-letter pronounceable .com inventory trades like small real estate in good zip codes — limited stock, patient holders, end users who show up with wire instructions. Two-word SaaS brandables on .com or .ai compress a category story into one string. Those don't replicate when Verisign prints another quarterly chart.

New gTLD growth adds experiments, not substitutes. A founder can register something clever for thirty bucks and still lose the board meeting to a competitor on a clean .com. I watched that play out twice this quarter in robotics — descriptive .com beats coined nonsense every time a procurement team gets involved. Names like CobotHand.com sit in that lane: obvious out loud, credible on a pitch deck, no lab-notebook spelling tax.

Meanwhile average brandables sit longer even in warm markets. I wrote about patience pricing in why brandables sit for months — macro registration growth doesn't shorten the funnel if your name is abstract syllable soup. Buyers have more cheap options at the bottom. They don't have more great options at the top.

That's the investor edge, if you still want one in 2026. Own the upgrade path — the name they'll reach for after the twelve-dollar compromise embarrasses them in a sales demo.

What should domain buyers actually do with Q2 2026 data?

Stop treating headline totals as a price forecast. Track closed sales, extension-specific velocity, and your own sell-through — not the global counter. Mid-year dollar volume was already running hot before this DNIB landed; our midyear 2026 domain sales breakdown showed .com and .ai still clearing serious checks while disposable inventory churned underneath.

Sellers use DNIB graphics as mood lighting. Buyers should use weekly charts on DNJournal as comps anchors. Registration counts measure supply entering the stadium. DNJournal and NameBio measure who paid retail to leave with the good seats.

If your thesis is "I'll wait until we hit 450 million and prices collapse," you're betting against fifteen years of data. The gap between "a domain exists" and "the right domain is available" keeps widening. I'd rather be early on one sharp name than late on ten maybes.

What I'd do this week as a buyer or portfolio holder

  1. Download Verisign's Q2 DNIB and highlight which TLD classes drove the +9.1M — separate world totals from the English-brandable market you actually trade.
  2. Pull ninety-day comps on NameBio for five names you'd buy today; note median sale price, not the outlier that makes Twitter loud.
  3. Run a renewal audit — kill anything you wouldn't repurchase at ask if you saw it fresh; 400 million domain names means more junk competes for attention, not more trophy supply.
  4. Shortlist three category-forward .com or .ai names with obvious end-user stories; industrial, SaaS, and AI verticals still show upgrade hunger.
  5. Check WHOIS and trademark lanes on your keepers using domain tools before outbound season picks up.
  6. Re-read buyer FAQ escrow steps so you're not improvising when a July DNIB headline makes a seller greedy.
  7. Make one offer on a name you'd proudly attach to a product — not a lottery ticket you'd hide from your spouse.

That's the boring playbook. It works when macro numbers get loud.

How I'm thinking about my own portfolio right now

Fewer names, higher conviction. I'm overweight descriptive-but-brandable .com where buyers default-trust the extension — robotics, manufacturing, B2B tooling. I'm underweight long-tail keyword pairs that only made sense when registration counts were lower and SEO arbitrage was easier.

Renewal math gets ugly on a 2,000-name junk portfolio. At 401.6 million total registrations, the penalty for holding mediocre inventory is higher because end users can always register "good enough" for twelve bucks. Your listing needs to be obviously better — shorter, cleaner, more category-native. I cut twenty-three names last month. Felt good. Should've cut forty.

I'll browse curated inventory before I hand-reg another hyphenated compromise. Not in the opening minutes — after the comps work. When I'm ready to buy operator-grade names, I start at the premium domain marketplace because landers with positioning beat raw WHOIS hunting on a million-result grep.

Second look at CobotHand: if you're shipping cobot hardware or end-effector software and you've been burning months on coined brands nobody spells after one trade show, CobotHand.com is the kind of string that ends the naming meeting in one slide. Industrial categories punish cute. Ask me how I know.

I'm not selling fear of the milestone. I'm selling focus. The internet will hit 450 million. Your shortlist shouldn't grow to match.

I'll watch Q3 the same way I always do — spreadsheet first, narrative second. If you're waiting for 401.6 million domain names to crash trophy pricing, you might wait forever while the names you actually want keep climbing. Make the shortlist. Run the comps. Send the offer.