For about a decade now, somebody has been writing the obituary of .com. The argument repeats: there are hundreds of new TLDs, the good .com names are all taken, “users don’t care anymore.” The data does not back this up. .com is still where premium domain spend concentrates, by a wide margin, and there are specific reasons the moat hasn’t filled in.
There are also specific cases where insisting on .com is the wrong call. This piece covers both.
The three reasons .com keeps its premium
1. Inheritance, not preference
When users hear “the website is widget”, their fingers type widget.com reflexively. This is not a preference, it’s muscle memory built over thirty years of .com being the first thing taught about the web. You can change muscle memory with massive marketing spend (https://x.com/’s rebrand from .com to .com is hilariously circular but you take the point). For a startup without that budget, the inherited reflex is a real conversion factor.
2. Email trust
The deliverability factor is almost always under-discussed. Outbound email from a .com address has measurably better inbox-placement rates at major receivers (Gmail, Microsoft 365, Apple Mail) than the same email from a less-established TLD. The receivers’ filters use TLD reputation as one of many features, and .com has the longest history of legitimate sending behind it.
For a B2B SaaS sending hundreds of emails per day, this can be the difference between cold-outbound landing in an inbox or a spam folder. Getting it right at registration time costs an extra few thousand dollars; getting it wrong costs you customers you never knew you missed.
3. Resale value floor
Premium .com names hold value across decades. A premium .io from 2018 is worth roughly the same in 2026; a premium .com from 2018 is generally worth more in 2026, sometimes much more. If you ever want to sell the brand, sell the company, or recover the domain investment, .com is the most liquid asset class.
This isn’t a guarantee — bad .com names do depreciate — but the floor for genuinely good ones is sticky.
When the alternative is the smarter pick
When the .com is unobtainable at any reasonable price
If [your-name].com is owned by a multinational who won’t sell, or by a domain investor asking $250k when your seed round is $1M, the right choice is often a high-quality .io, .co, or .ai. Spending 25% of your seed on a domain is unusual; spending 5% on the next-best TLD and putting the difference into engineering is usually better.
When your audience is unambiguously technical
For a developer-tools or AI-infrastructure product, .io and .ai carry no penalty among the buying audience. GitHub, Stripe, Vercel, Cloudflare — all-.com, but the next tier of credible dev companies live perfectly fine on .io. If your buyer types linear.app without thinking twice, you don’t need to fight for the .com.
When the .com is taken by someone semantically wrong
If widget.com is a paint company and you’re a SaaS, the risk of confusion at the address bar is mild (different industries, different muscle memory clusters). Picking widget.io and using it consistently is fine — and may be cleaner than buying a .com you’d then have to defend trademark-wise.
When you can afford a non-matching but distinctive name
Some of the best brands of the last decade chose distinctive names and .coms that weren’t exact matches: Stripe started on stripe.com only after an acquisition; Notion was notion.so for years. A short, distinctive, easy-to-spell name on any reasonable TLD beats a long, awkward .com.
Our actual heuristic
When we advise on a brand-naming decision (and we do, when buyers ask), the rough hierarchy is:
- Premium short
.comof a real word or invented brandable: best - Premium short
.comof a compound: very good - Premium short
.io/.ai/.appfor technical brands where it’s the natural TLD: good - Premium short
.cofor consumer / lifestyle brands: good - Long
.comwith a forgettable or hard-to-spell name: worse than any of the above - Long compound on a non-
.com: usually bad — the worst of both worlds
The cost difference between row 1 and row 6 can be 100×. The brand outcome difference, over five years, is also usually 100×.
.com still wins, but the reason is mechanical (muscle memory, email reputation, resale liquidity), not magical. When the mechanics don’t apply, neither does the premium.