There's a name I picked up at auction in 2019 that I paid far too much for. My walk-away number that afternoon was $1,200. I got stubborn around eleven at night against a bidder I'll never meet and closed at something like $3,400, and the next morning I told myself I'd been decisive. Two years later an offer came in at $2,600 and I didn't even counter it, because $2,600 was a loss and I wasn't in the mood to book a loss that week.
None of the auction had anything to do with the domain. It was a fact about me on a Tuesday night. I let it set the asking price for three years anyway.
The cheap version of the same mistake is easier to miss because it feels like winning. I sold a hand registered name for $900 once, maybe six hours after the first email came in, and only found out afterward that the buyer had filed a trademark on the term two months earlier. Nine hundred dollars looked like free money to me because I had eleven bucks in the name. To him it was a rounding error on a launch he'd already committed to.
The buyer doesn't know your cost and wouldn't reprice if you emailed him the invoice. He's weighing your name against whatever else his shortlist turned up, and against how much of the budget his co-founder will sign off on without a conversation about it.
So take two names of roughly equal quality. One investor caught his in a drop auction for $500. The other chased the owner for a year and finally paid $8,400 in a private deal. If the comps put both of them around twelve grand retail, they're both twelve grand names. The second guy didn't buy a better domain. He bought the same domain badly, which happens to all of us and isn't something the market is going to reimburse.
The reverse is where I think most of the money actually gets left behind. Say you registered something five years ago on a whim for ten dollars and the term caught on since then. There are companies using it in their marketing now, a few close comps have sold in the mid four figures, and you could sit down and list six or seven businesses that would be better off owning it than you are. You've got maybe sixty dollars in it after renewals, and that number is going up a little, since Verisign is raising the .com wholesale price again in November. Sixty dollars is not a valuation input. All it tells you is that your carry has been cheap, which is a nice thing to know and completely useless for pricing.
The psychology part is hard to write about honestly because everybody nods along at sunk cost and then goes right back to doing it. Selling at $3,000 on a name you paid $5,000 for feels like writing somebody a check for two grand, so you hold at $7,500 and call it patience. Sometimes it really is patience and the buyer shows up eighteen months later. More often you're protecting yourself from the moment you have to update the spreadsheet and admit the buy was bad.
Cost basis does matter, just not where people keep putting it. It belongs in your return math, in how much capital you're willing to leave parked in a single name, and in whether you can afford to hold out when an offer lands, because $25,000 tied up in a domain buys a very different kind of stubbornness than $500 does. Those are portfolio decisions. They have nothing to do with what the name is worth to the guy who just emailed you.
When I get stuck on a price now I try to ask what I'd bid if the name came up at auction tomorrow and I'd never laid eyes on it. It isn't a clean test. You can't really un-know what you paid, and I'm not sure anybody has ever answered that question with a completely straight face. But it beats the receipt, and it usually pushes me back into doing the work instead of defending a number. Pull the comps. Write down who the buyers plausibly are. Look at what those buyers could go register instead of paying you. Reread the last few inquiries and notice how fast they went quiet and at what number.
The 2019 name eventually sold. I lost money on it, not a devastating amount, and the sale came maybe four months after I quit pricing it like the auction had meant something. I don't know that those two things are connected. Could have been timing.
What I do know is that I have three or four names in the portfolio right now that I'm still pricing off what I paid for them. I can watch myself doing it. I haven't fixed it.