Service words vs short strings in the .si market

Most .si pricing advice reduces to a single variable: how short the name is. We think that is the wrong axis. Across the 234 tracked .si sales in our database, the split that survives contact with the numbers is not short against long — it is a word a business needs against a string somebody liked. The two overlap, but they are not the same thing, and the overlap is where most of the confusion lives.

What the tracked set is

These are venue records and disclosed sales, not an estimate model. The set holds 234 sales, with a median of $3,599, an average of $5,085 and a largest sale of $50,000 for staffing.si through Spaceship. Every number below comes from that set, the one you can filter yourself on the sales index.

The measured shape, by length

Here is the length table exactly as it stands:

Characters	Sales	Average
2	12	$6,736
3	11	$4,397
4	37	$4,769
5	47	$4,774
6	35	$4,386
7	30	$5,438
8	26	$5,879
9	17	$6,258
10	6	$7,050
11–13	13	$3,201

Read across it and there are three behaviours, not one slope:

• Two characters — 12 sales, $6,736 average. The highest average in the table, on the smallest base, and the lowest supply by construction.
• Four to six characters — 119 sales, about $4,658 average. Flat and unremarkable: the largest group in the set and the closest thing to a market baseline.
• Seven to ten characters — 79 sales, about $5,882 average. Higher again, and this is the band where service and category words sit.
• Eleven to thirteen characters — 13 sales, about $3,201 average. The one place the average falls away sharply, but 13 sales is not a sample; it is an anecdote we are telling you about.

So the familiar claim, that value falls as length grows, is not what the tracked set shows. Price dips through the crowded middle and recovers where a name starts describing a business.

Where the service words actually are

Four of the ten largest tracked .si sales are words that describe what a business sells or serves:

• staffing.si — $50,000 (Spaceship). A service word, eight characters, and the largest sale on record.
• gifts.si — $31,346 (Fruits). A category word, five characters.
• stocks.si — $25,000 (Sedo). Category, five characters.
• healthy.si — $23,000 (Private). Category, with obvious clinic and retail use, seven characters.

The rest of the top ten — hollywood.si, runtime.si, pure.si, caspian.si, dubai.si, plasma.si — are place, brand and technical words. They are not service words, and they are not short either. Since gifts and stocks are both five characters, this is not a length story at the top of the table. It is a commercial-intent story with length as a rough correlate.

Why the two ends behave differently

Two unrelated scarcities push prices up, and the middle has neither.

• Short strings are scarce as inventory. Our records contain 12 two-character sales in total. A buyer competing for the two or three that suit them cannot substitute around the shortage, and holders know it.
• Service words are scarce as meaning. Excavators, chauffeurs, petsitting, glassrepair — the words a local business is effectively named after — are wanted by essentially one buyer: the person running that business. There is no substitute name and no substitute buyer.
• Four to six characters is the crowded middle. Too long to inherit the scarcity premium of two letters, too short to carry a service description. That is why a $4,386 average for six characters sits below both ends of the table.

Length is a proxy for two unrelated effects that happen to push in the same direction at opposite ends of the scale.

Where this evidence runs out

We would rather flag the gaps than paper over them.

• We hold no category field. Each record carries a name, a price, a venue and a date. "Service word" is our reading of a name, not a field the registry or the venues publish. Anyone else could classify the same set differently and reach a different split.
• The example words have no comparables here. excavators.si, chauffeurs.si, petsitting.si and glassrepair.si do not appear as tracked sales at all. We can argue from the shape of the table; we cannot price them from receipts we do not have.
• The eight-character average is largely one sale. $5,879 across 26 sales includes the $50,000 outlier. Remove it and the remaining 25 sales average about $4,114. Most of the apparent advantage of eight characters is a single transaction.
• The long tail is very small. Everything from eleven characters up is 13 sales in total, so treat any pattern there as a hypothesis rather than a finding.
• The whole market is thin. 234 sales is a census of what was disclosed, not a liquid market. On a base this size, one deal moves a bucket average by hundreds of dollars.

What the sample does support is a mild, directional claim: descriptive words in the seven-to-ten character band average above the four-to-six band, and a service word holds the outright record. What it does not support is a formula, or a price for a name nobody has sold yet.

What we do with it instead

Because the set cannot carry a formula, we do not present one. Any name entered on the appraisal page returns a range with its comparables, its arithmetic and a confidence level attached, and the published method shows which signals are doing the work. When comparables are too few to separate two words, confidence drops and the page says so — the honest answer for a market where the example words above have never sold in public.

For a seller, the practical reading is narrow: if the name is a service or category word, do not discount it for being long; if it is a short string with no meaning beyond its scarcity, price it against a two-letter base of twelve sales and expect to hold it. Either way, whether the obvious name you want is still available is a question the registry check on our home page answers in a second — free, and with no account.