Joseph Peterson reviews recent NameJet sales, and takes a deep dive into sales of 3 and 4-character domain names.
After a record-setting Winter, NameJet began Spring with a reduced sales volume for March. Even so, given 94 sales closing at or above $2000, March matches the best month out of an 11-month stretch during 2014. NameJet also scored its 4th highest sale of all time with VVV.com being bid up to $136,400.
The real story here is the massive preponderance of 3 and 4-character domains. Out of $641.6k in sales, $306.1k (47.7%) derive from CCC domains with another $115.1k (17.9%) from CCCC domains. That leaves only 34.4% – basically 1/3 of high-end spending at NameJet – to be split amongst ALL other domain categories. Looked at another way, there were 14 CCC and 21 CCCC sales, which together account for 35 out of 94 domains in our chart: 37.2%. In other words, 1/3 of Namejet’s best-selling domains were 3-4 characters long and accounted for 2/3 of high-end spending. Out of the top 21 sales during March, the only slots not occupied by these 3-4 character strings were #13, #15, #18, and #19. Keep in mind, we’re counting only domains that cleared $2k individually. Below that high-visibility threshold, the domain profile will be different.
All of this boils down to a single word: China. Not all buyers were Chinese, but even those that weren’t will mainly sell their short domains (directly or indirectly) into Chinese hands. And the prices they pay sometimes indicate a borrowed mindset – one that buys into market values ultimately set by China. Such domain traders are merchants on a digital silk road connecting East and West. And as long as the Chinese appetite for numeric / literal strings remains strong … or is perceived to be strong enough to keep these assets liquid outside China, just so long will domainers from Denver to New Delhi be found bidding $14k – $20k for the likes of SZP.com, 1303.com and NQJ.com – arbitrary and alien as such commodities seem to us. Even though I don’t think a downward correction in short domain prices is imminent by any means, it does bear thinking what NameJet (and domainers generally) would do if China decided tomorrow that LLLs and NNNNs were passé.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
VVV.com | 136,400 | NLF.com | 33,700 |
QMS.com | 22,342 | SZP.com | 18,109 |
PZM.com | 17,008 | JZC.com | 16,910 |
NQJ.com | 15,619 | VOJ.com | 11,312 |
LFV.com | 11,100 | GVY.com | 9200 |
VVV.com is notable because it follows the December sale of its cousin, VVVV.com. I’m sure the seller wished to capitalize on the success and publicity surrounding that earlier NameJet auction. Clearly the strategy paid off. As for the others, well, they’re somewhat (though not exactly) interchangeable. The category is fairly boring – like buying lumps of metal by the ounce, based on measurable purity. Frankly, it’s because one LLL.com is so like another that the asset class is as liquid as it is. Buyers from all linguistic backgrounds and experience levels feel somewhat confident in their valuation because there’s little to evaluate in any individual specimen, while (at the same time) there is an enormous pool of equivalent assets to compare.
Sales pitches for LLL.com domains make much of their scarcity (only 26^3 = 17,576 in existence … and many of those off the market), but ironically the high prices of LLL.com domains are really based not on their scarcity but on their abundance! Domains grounded in words rather than letters are each so semantically unique that global buyers have trouble assessing and comparing them. It’s simply impossible for the market to generate as much consensus about the value of Ankles.com ($6.1k) as observers can instantly assign to something as meaningless to them as PZM.com. For PZM.com has thousands of nearly identical peers, whereas Ankles.com may have (at best) a handful; and even there comparisons are subjective.
The LLL.com boom isn’t entirely due to China. I’d say that it’s an American bubble that would burst except for China’s growth. Uncertainty about how to appraise linguistically meaningful domains – not to mention the difficulty in marketing them for sale to buyers who are equally unsure of their worth – depresses the value of items such as Ankles.com relative to fast-trading, stable commodities like PZM.com. Market consensus reduces risk; so in times of insecurity (think nTLDs) or poor wholesale liquidity, money flies to the most secure asset class, driving prices upward as a symptom of low-level mass anxiety. And in the case of the domain industry, we’re seeing that happen now.
LLL.com traders are packed tight as sardines; and they pass their assets around so frequently that the niche conducts “current” very well. Indeed, LLL.com domains behave more like stock market shares, barely resembling the rest of the domain market, where branding and business models, search stats and online competition must all be taken into account for each and every domain as an individual. Not so with LLL.com’s. Most domain investors by now regard them mainly as abstract place holders for money. Margins can be quite tight – narrower than in most domain asset classes. At every trade, brokers and market places exact a commission; so understandably LLL.com’s are also highly promoted because they sell fast and sell high. Frequency of trades means that values are nearly unanimous, plus or minus. I’ve long thought that it’s something of a self-fulfilling prophecy. Notice, LLL.com price ranges are so well established that none sold below $9.2k. As long as everybody agrees that the floor price of LLL.com domains is X, then everybody pays X. But why X? Because it was X yesterday – not for any intrinsic reason. Profit depends on a change of venue – a visibility advantage this seller has over the last seller. Or else profit depends on waiting for the floor price to rise … perhaps driving it to rise by publicizing sales.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
1303.com | 14545 | 8316.com | 13345 |
7610.com | 12601 | 3202.com | 11000 |
6105.com | 9000 | 6474.com | 7701 |
Numerics, even more so than literal strings, are a domain category dominated by China. After all, NLF.com ($33.7k) has some value as an acronym anywhere the Roman alphabet is used. So where quality acronyms are at issue, there’s often a tug-of-war between China and institutions in the western hemisphere. But using numbers to abbreviate words and phrases may be a practice unique to China. As you can see, the price range on these half dozen 4-digit .COMs is rather narrow: $7.7k – $14.5k with nothing below $7.7k to report. For another domain asset class – say, dictionary words – prices will always run the full spectrum down to zero.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
000001.com | 3500 | 860888.com | 3101 |
15200.com | 2111 |
Plenty of domain investors have heard by now that the digit “8” is thought to be lucky. Repetition also increases value for Chinese buyers. Normally, of course, we wouldn’t see 6-digit domains outperform 5-digit domains. 000001.com is not a first, but it is an oddity.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
M86.com | 3400 | IZ8.com | 2901 |
WT8.com | 2622 |
Even though China prizes short strings enough that bidding activity on mixed alphanumerics is largely attributable to the Chinese market (at some stage in the supply chain), still … not every domain China likes China gets. M86.com might be a cyber-security company, as I reported earlier.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
HOBB.com | 7777 | KYTE.com | 5303 |
RRYY.com | 4200 | SUZA.com | 2985 |
VVTT.com | 2902 | GSVC.com | 2701 |
AUTU.com | 2600 | LLOO.com | 2545 |
BAPY.com | 2500 | YENS.com | 2450 |
TCAS.com | 2410 | EATE.com | 2222 |
WDXX.com | 2149 | BONZ.com | 2109 |
HALY.com | 2109 |
Mainly what we see with top-selling LLLL.com domains is some combination of these factors: (1) word resemblance, (2) pronounceability, (3) repeating letters, (4) a CVCV pattern favored by many domain investors, and (5) applicability as an acronym. You can judge for yourself which factors apply to which domains above. Interestingly, pronounceability and usefulness as an acronym are often vitiated by repeating letters. Pattern is most important among Chinese buyers, for whom our criteria don’t matter much. Some of these domains I’ve covered in more detail within my weekly articles on the expired market: HOBB.com, KYTE.com, and BONZ.com.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
BaoCan.com | 5137 | TaoKuang.com | 4559 |
YaoBi.com | 3209 | LaoHu.net | 5020 |
China also pays good money for phonetically spelled Pinyin. Usually that means .COM. Yet here we have NameJet’s only .NET to clear $2k during March. Another Chinese-style .NET sold during the same week; but technically that spilled over into April.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
Alibab.com | 5200 | Propero.com | 4200 |
We’re not done talking about China yet. I’ve paired these domains because they both seem to apply to preexisting brands. If you don’t recognize the company Alibaba.com in that typo, then you can read a bit more about it here, along with Propero.com.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
Shawls.com | 6900 | Ankles.com | 6099 |
Negatives.com | 3000 | Punches.com | 2800 |
Fondly.com | 2500 |
At long last, I can finally discuss domains built from intelligible words rather than character sequences! Here we have 5 single words from the English dictionary, which are for me at this point like water from an oasis. Strangely, the top 4 out of 5 are all plural nouns. Shawls.com ($6.9k) can be shopped & shipped; Ankles.com ($6.1k) can be sprained & treated; Negatives.com ($3k) can be developed as photographs; Punches.com ($2.8k) can be rolled with or brand anything that “floats like a butterfly, stings like a bee”. Fondly.com ($2.5k) should do well, since startups are exorbitantly fond of this adverbial “-ly” suffix. Personally, I hate the fad. But this name is short, positive, and genuine (as opposed to manufactured). Plus, the internet these days is all about liking things – i.e. expressing fondness. So I’m as fond of Fondly.com as the buyer must be.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
PrepSchools.com | 8888 | BearClaw.com | 8600 |
BabyPhoto.com | 4944 | JetRental.com | 4700 |
SleepStudy.com | 4300 | MailMarketing.com | 4210 |
ElectricHeaters.com | 3800 | InternetRadios.com | 2602 |
NestingDolls.com | 2009 | RingBands.com | 2000 |
It’s easy enough to conjoin pairs of words. But these 2-word domains each define 1 real thing, and the phrases are really used to describe that thing. Neither word is extraneous – meaning that if you erase “prep” from PrepSchools.com ($8.9k) or “ring” from RingBands.com ($2k) you actually change the topic entirely. Under those conditions, 2-word domains can do quite well on the market. I’ve written about some of these auctions already, including 2 with real profit potential. Where MailMarketing.com is concerned, Andrew Allemann has written a couple of interesting pieces lately reflecting on TLD usage within coupon mailers he has received.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
MyBooks.com | 11100 | LuckyGames.com | 9501 |
TransNet.com | 7499 | FancyCars.com | 3500 |
WebSurveys.com | 2400 | InfoShield.com | 2269 |
AllergyAware.com | 2100 |
Unlike the last group of 2-word domains, these phrases are more optional than actual. That doesn’t mean that such names don’t have value – only that creativity and subjective taste play a large role that can be hard to measure. Knowing the scale of the underlying industry really helps predict what kind of prices companies can pay for a keyword-based but “brandable” domain. Notice that these auction results all apply to big industries: books, gambling, transport, translation, automobiles, surveys, cyber-security / privacy, pharmaceuticals. What I’m getting at here (for the benefit of any new domainers who may be paying attention) is that, while brainstorming creative names about nesting dolls may be fun, such a tiny industry won’t pay much for domains in this style so long as it can afford to buy NestingDolls.com itself for $2k. So it’s best to pursue domain investment in the more creative names only for the bigger industries.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
0Down.com | 4503 | SwapStuff.com | 3388 |
SecretLover.com | 2600 | LocalShops.com | 2588 |
SoftwareDeals.com | 2200 | PhoneSavings.com | 2100 |
WeRoll.com | 2000 |
These 2-word phrases may not define actual things, but they are commonly said or else seem perfectly natural. Moreover, the first 6 clearly apply to lucrative niches: loans, bartering, adult dating, local directories / shopping, software, and phone plans. That means that people looking to brand a business or website within these niches will frequently come up with these very same domains as name candidates. Greater market demand means greater prices … plus greater likelihood of a sale at any given price. I particularly like WeRoll.com ($2k) and SwapStuff.com ($3.4k), both of which I reviewed in earlier articles.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
Epitech.com | 2988 | Finsol.com | 2311 |
Frequently, when domainers discuss “brandable” domains, they may be referring to neologisms such as these – invented pseudo-words that may or may not contain parts of real terms but which inevitably hint at vocabulary and language origin. Epitech.com matches the brand name of a few developed sites, and perhaps Finsol.com does too (I haven’t checked). In that case, I might as easily have grouped these with Propero.com and Alibab.com above, since the opportunity to protect a brand or flip to an established company largely explains the auction prices in such circumstances. Wholesale prices tend to be far lower otherwise, which is why we see few neologisms at NameJet above $2k … unless they match one or more predetermined buyers.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
SealBeach RealEstate.com |
3866 | ChicagoDivorce.com | 2755 |
PembrokePines RealEstate.com |
2621 | WeBuy HousesCash.com |
2600 |
DiscoverTexas.com | 2422 |
Domains about real estate frequently do well – even though a geographical component invariably makes them longer than other domains in the same price bracket. And the reason is simple: the products they sell (houses and other property) are among the most expensive. Adding a geographical term also makes sense for locally circumscribed professionals such as lawyers (ChicagoDivorce.com). The same obviously applies to tourist destinations (DiscoverTexas.com).
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
USD.org | 5450 | OnlineBackup.org | 3188 |
Voices.org | 2801 | TurboTap.org | 2322 |
DebitCard.org | 2225 | Helix.org | 2099 |
Out of 94 domains that sold over $2k during March at NameJet, 87 were .COM, 1 was a .NET in Chinese pinyin, and the remaining half dozen were .ORG. Helix.org ($2.1k) relates to DNA. The other 2 single-word domains are self-explanatory. Among the trio of 2-worders, TurboTap.org is the odd man out. Unlike DebitCard.org and OnlineBackup.org, it doesn’t define any commercial category. Instead, it looks to be an expensive bit of brand protection. At first, I suspected TurboTax; but “X” and “P” are too far away on the keyboard. Still haven’t fully researched that one, though; so I could be entirely wrong. Although the .ORG that fetched the highest price was an LLL, in keeping with the majority NameJet’s highest .COM sales during this same month, the rationale is clearly different. We haven’t seen China throwing money at .ORG strings, and “USD” is far from an arbitrary sequence.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
Banerjee.com | 6901 | Beston.com | 2421 |
BWIN9.com | 2324 |
These are surnames. So is HALY.com from the table of LLLL.com domains above. BWIN9.com seems to exclusively match one man’s user name at various sites. Yet he wasn’t the buyer; someone in China was. Perhaps the word “win” indicates gambling. Still something of an enigma, this one; but I had to stash it somewhere.
Domain Name | End $ | Domain Name | End $ |
---|---|---|---|
Makanan.com | 3363 | Servizi.com | 3020 |
Lirio.com | 2615 | Eletro.com | 2151 |
Inizio.com | 2101 | Draken.com | 2099 |
Not everything in the world (or at NameJet) divides between English and Chinese. Plenty of domains in other languages sell each week, although the wholesale prices at NameJet tend to fall below this $2k monthly reporting threshold. Makanan.com is Indonesian for “food”. Servizi.com is Italian for “services”. Lirio.com ($2.6k) is Spanish for “lily” or “iris”. Eletro.com is Portuguese for “electric” or “electronic(s)”. Inizio.com is Italian for “beginning” or “I start”, as well as being used as a brand name by several American initiatives. Draken.com (I’m told) means “dragons” in Dutch; and it’s “dragon” in Swedish too, which explains why the Swedes manufactured actual flying Draken.
I’ve sold perhaps 15 3l .coms to Chinese buyers recently (last month). The floor price is about $20k and as high as $38k. Some buyers have bought up to 3 names. These aren’t packages but individual sales. Buyers are coming in at about 10k to 12k but seem to be willing to pay much more than they did 6 months or 1 year ago. There is nothing special about these 3l’s either. Just random letters and some have “q” “x” and so on as well. Most of these are purchased direct (from us) but some are deals that brokers in China have brought to us (with their markup). There appears to be more of an active trade at realistic pricing than there ever was at least that I can remember. Judging by multiple bids (on the same names) it appears that they are somewhat liquid and a way for Chinese buyers to get money out of the country which I believe is part of what they are trying to do (same with buying US real estate in hyped up markets like NYC).
Getting money outside the country is probably a motivating factor for Chinese investors. They’re buying property / tokens of all kinds. So I think you’re right there.
Prices on LLL.com, CCC.com, and NNNN.com domains are tied to the Chinese economy and mass investor behavior in a way that doesn’t apply to the majority of domain names. They’re predictable enough that buy / sell decisions could be automated. Algorithmic traders could step in and clean up. But they’re vulnerable to downturns or spending shifts within China, even if the general upward trend has been quite stable.
Really, this is a disjunct market. Other domain asset classes depend on larger decisions by a smaller number of buyers – niche end users branding a website … or speculators weighing the odds concerning such end users and their budgets and subjectively assessing the merits of a given name. Far more research, uncertainty, and marketing labor.
We’re looking at 2 distinct domain markets. They overlap in the sense that they share venues to move inventory; and money sloshes between them, since owners and brokers often hold both word-based and character-based domains. But I’d say that in terms of market behavior these 2 asset classes are different enough to be classified as separate species.
Extremely valuable analysis. Appreciate all the work that must have gone into this one. I’m really intrigued by China’s entry into the LLL.com and numerics markets… I wonder if a recession in China – such as the one that hit the USA and Europe in 2008 – could impact these types of domains? Or will these assets forever appreciate in value?
I don’t know.
But I’d ask myself this question: Are short-string .COMs behaving like property or like tokens?
During the 17th century, the Dutch went crazy for tulip bulbs. At first they were traded as property, desirable for their own sake and somewhat rare. The Dutch even evolved a kind of futures market for tulips. But speculation drove up prices until tulips were behaving more like tokens than property – receptacles for monetary value.
Tulips happened to be a bubble that burst thanks to the black plague. But plenty of abstract tokens are traded in a very stable fashion. Gold, for example, has been a token for millennia. Even if the metal ceased to be used for jewelry and had no use in manufacturing, it would continue to be traded as an abstract token for value, based on the stability / inertia of the market itself.
At a certain point, when an asset is worth more as a token than it is as actual property, then it becomes unusable. You wouldn’t plant your tulip bulb if it’s worth as much as your house. You wouldn’t call in your certificates for gold in order to melt it down and plate a statue of yourself, since that reduces the liquidity of the token.
So if LLL.com, CCC.com, and NNNN.com domains begin to behave more like tokens than like property, Chinese investors may reach a point where they’re so invested in their hoard of short domains that they refuse to develop them. After all, if the business model for a website can only justify an expenditure of X for its brand name / domain but the LLL.com is trading purely as a token for multiples of X, then it would be a waste to take that token off the market and bury it in a website. Instead, a Chinese investor would maintain the token ready for liquidation, unencumbered by any actual usage, and pick some other type of domain to function online.
The value of domains as property – i.e. as digital land to be built upon or as brand names to be selected by online projects – cannot appreciate indefinitely. A plumber or even a well-funded social app will have real budgetary constraints that limit what any domain can be worth to them.
But the value of tokens is arbitrary. Maybe we’ll wake up in 2020 to find that planet earth universally recognizes LLL.com domains as being worth precisely $250,000 each. The margin could be that tight and the value that high. There’s nothing to rule that out in principle, if LLL.com domains behave like tokens. They’d simply be equivalent to quarter-million-dollar bills. And as long as everyone within the economy unanimously agrees that these tokens are worth $250k, then they’ll function as other abstract representations of money function – like bills and bonds.
Really, it depends on the disconnect between property / usage and abstract trading. If short-string .COMs take flight and leave the earth behind, then they can be worth anything people imagine and agree that they’re worth. But in that case, they’d begin to vanish as websites, simply existing as cash-receptacles within investor portfolios.
Right now the market can’t quite make up its mind whether such domains should be bound by their value as property, something to be put to use by websites, or whether they’re better off being traded like tulip bulbs one owns but never touches. Maybe gravity will apply; maybe it won’t, and LLL.com prices will float up and away like balloons set free.
Wow! Absolutely fascinating perspective. I’ve been mulling over this for a few hours now. Much appreciated, Joseph!
Hi Joseph thanks for an excellent article and analysis. It’s very useful to separate domains into two markets, usable goods and tokens, but I’d like to pick up on your point in the comments about not using an LLL for a site once it has become a token. The investment in the site is not sunk into the domain: you can continue the site on a replacement domain, albeit losing any links built up unless the old domain is set to forward to the new. So the token status of an LLL does not preclude use for a site, but I agree it probably makes it less likely.
In China, a lot of people has emails which contain numbers. For a lot of them the numbers are actually their phone numbers. It makes easier to be shared with people outside of their country. It is a smart thing! I guess that’s why they buy also numeric domains, to be easier for them to share.