We sorted the 139 .com names on our own shelf by asking price: 27 names carry 36.4% of the money, 85 carry 56.1%, and 27 carry 7.5% with a $375 median. How to build the bands, what each one is for, and the four decisions that follow.
A portfolio audit sounds like a spreadsheet exercise and is really a decision exercise. The question is never "how many names do I own". It is "which of these names pays for the other ones, and for how long am I willing to fund that".
We ran the audit on our own shelf this week: 139 .com names with published buy-now prices, sorted by asking price, cut into three bands of roughly twenty, sixty and twenty percent of the count. The result is not a bell curve and it is not the "80/20" every domainer repeats. It is a shape with a specific set of problems in each band, and the decisions are different in each one.
What the audit actually measured
Three numbers first, because an audit is only as good as the column you sort by.
- Count of names with a published buy-now price: 139 .com, alongside 44 .si and 7 .net on the same shelf.
- Total asking value of those 139 names: $141,704.
- Median ask: $940. Lowest: $195. Highest: $2,800.
One caution before any of it: an asking price is not revenue. Nothing here says a name will sell, or that the shelf has ever transacted at these numbers. That is the audit's problem, and we will get to it. The prices are our current published asks; the count and the spread are the only two things you can trust at this stage.
We then sorted descending by price and split the list three ways: the top 27 names (about 20% of the count), the middle 85 (about 60%), and the bottom 27 (about 20%).
The three bands, with the numbers
The top band: 27 names, $51,634 of asking value, 36.4% of the shelf. Median ask $1,870.
That is not 80% of the money, which is the story people expect. It is just over a third, concentrated in a little under thirty strings. These are the names that carry a single-name story: a two-word .com in a commercial category, a short brandable, a service noun with a buyer class behind it.
The middle band: 85 names, $79,509, 56.1%. Median ask $940.
This is where most of the asking value actually sits, and it is the band nobody writes about. Sixty names would be 43% of the count; here 85 of 139 names is 61%, and they hold the single largest block of paper value on the shelf.
The bottom band: 27 names, $10,561, 7.5%. Median ask $375.
Twenty-seven names, a thirteenth of the shelf's asking value, none of them above $500 by construction. Two sit under $200. These are the names that renew forever and produce a conversation once a year at most.
So the honest shape of this shelf is: a third of the money in a fifth of the names, over half of the money in the middle, and a long tail of chalk.
Why the middle is the hard band
The instinct is to work the top band — those are the names where one sale changes the month. In practice the top band is the least work per dollar: 27 names, a clear story each, and a buyer who arrives already qualified by the price.
The bottom band is easy in a different way. Nobody expects much from it, and the correct move is usually structural: bundle it, auction it, or let it lapse.
The middle is the hard band because each name in it looks negotiable in both directions. A $940 name can become a $1,400 name with better copy, or a $590 name with a better listing, and you cannot tell which from the inside. Sixty per cent of a portfolio sitting in an undecided state is not a portfolio of assets; it is a portfolio of decisions postponed.
Three signs we now look for in the middle band, all of them visible without a market:
- A price nobody chose. A name at $940 that was set by copying a neighbour's price, or by taking an appraisal and halving it twice, has no trade-off behind it. If you cannot say what the buyer is comparing it to, the number is not a price.
- A category with no named buyer. "Could be a SaaS" is not a buyer. The names in the middle band that have sold or almost sold in our own history all had one named type of operator attached to them, and most of the stalled ones had an adjective.
- A listing that has never been tested. In the middle band, a name that has never had a reach-out, never been listed on a second venue, and never appeared in a bundle is not proven dead — it is unmeasured.
Where the 20/60/20 idea helps and where it does not
The version of the rule that circulates says 20% of names sell, 60% pay for themselves, 20% are dead weight. As a diagnostic for time allocation it is useful. As a prediction about your shelf it is fiction, because the split depends on the prices you set, not on some property of domains in general.
The split on this shelf, measured by asking value, is 36 / 56 / 8 — nothing like 20/60/20. Measured by renewal cost it would be different again. Measured by the number of hours a name absorbed last quarter it would be different a third time.
That is the point. Run the audit by at least three measures, and do not assume they agree:
- By asking value — the band structure above. Tells you where the paper is.
- By renewal cost — usually flatter, because a .com renews at roughly $12–22 a year regardless of its price. This is the measure that tells you what the shelf costs to hold.
- By hours spent — the least measured and the most honest. A $375 name that has received four slides in a deck, two emails and a listing rewrite has cost more per dollar than the $1,870 name that got one line in the same deck.
The last one is uncomfortable enough that most audits skip it. A simple version: for each name, count how many times it changed price, listing copy or venue in the last twelve months. The names with the highest count and the lowest price are the ones quietly consuming the calendar.
The arithmetic that decides the long tail
The bottom band's real cost is not the price it did not earn. It is the renewal, multiplied by the number of years you will keep holding it because it is cheap to keep and awkward to kill.
Move the same arithmetic to the top band — a $1,870 name at a $22 renewal — and the break-even interval drops from seventeen years to eighty-five. On the $141,704 asking value of these 139 names the renewal bill at $22 sits near $3,058 a year, and the top band alone — $51,634 — covers it more than sixteen times over. The bottom band carries $10,561 and covers it three and a half times. That is the real description of the long tail: not that it is worthless, but that it is the part of the shelf the calendar can absorb.
That single comparison is the strongest argument for the audit, and the reason we now compute it for every name before renewal. It also explains why the middle band matters more than it feels like it should: those 85 names renew at the same rate as everything else, and their renewals are being paid for by a much smaller number of eventual sales.
Four decisions that come out of the audit
An audit that produces a chart has failed. These are the four decisions we make after running this one.
The top band gets fewer listings, not more. Twenty-seven names, a single story each, one venue where the buyer class actually browses, and no experimentation with the price for a quarter. Experimenting with the top band is expensive; a $1,870 name whose price moves four times in a year teaches the market that the number is improvised.
The middle band gets one test per name, and a written deadline. One named buyer segment, one listing rewrite, one reach-out sequence. If nothing happens by the deadline, the name moves down a band or into a bundle — deliberately, with a date in the file, rather than lingering for another year.
The bottom band gets a structural exit, not a discount. Bundling ten slow names into one lot, or running them through a low-reserve auction, gets them out of the renewal calendar in one action. Cutting a $375 name to $195 does not; it just makes a cheaper name that renews just as often.
Nothing moves band without a dated note. The reason audits fail in practice is not the analysis; it is that a name moves between bands twice and then nobody remembers which band it was in. Write the band, the date and the reason next to the name, and re-read it at renewal time.
What this audit does not tell you
It does not tell you what the shelf is worth. It tells you how the asking value is distributed and what each band costs you to hold. Those are different numbers, and confusing them is how a portfolio gets valued at the sum of its stickers.
It also cannot tell you which of the 27 top names will actually sell this year. Nothing in a price list can. What it can do is make sure the names most likely to sell are the ones with the cleanest story, the fewest unanswered questions and a price that has not been fidgeted with — and that the names least likely to sell are not quietly paying their rent out of a calendar you did not audit.
The whole exercise took one afternoon, three columns and a sort. The decisions it forced will take a quarter. That ratio is the argument for running it — the analysis is cheap, and the postponement it exposes is not.
Buying a name this week?
139 .com names on a published shelf, from $195 to $2,800, buy-now with escrow-protected transfer. See the portfolio · the contact page
Buying a name this week?
180 exact-match, brandable and geo domains — buy-now price, escrow-protected, same-day transfer. See the portfolio · the contact page
