The Unofficial and Generic Debate of Automated Domain Appraisals

[user-7256]

Automated domain appraisals are a controversial subject, so I'd like to host a debate regarding this topic. I would like to keep this generic (no advertising, affiliations, etc, but resources are fine) as possible to prevent ... unwanted commotion.

As you read the following keep in mind that the point of appraisals is to get another's OPINION on the value of the name. Much research can be done but each will value the name differently in most cases.

First, let me list some cons:

• Computers cannot sound out a name
• Certain extentions, especially .net and .info can be difficult to appraise because of a wide variety of uses
• Computers cannot comprehend the common meaning of a particular word or short phrase, it must use external sources

Now some pros:

• Automated appraisals can bring about a wide array of data in seconds (e.g. OVT, Google, Alexa rank, and more)
• You don't necessarily need to have the name registered to have it appraised
• You can get an answer in about 2 seconds

And finally, some undecided factors:

• Computers cannot possibly know the current market demand (or can they... by outsourcing possibly?)
• Computers cannot appraise 1, 2, 3, or 4 character names.
• How much information should the computer be told before the process is considered "pointless/worthless"? (Obviously the name, and probably traffic/revenue, but individual keywords and/or generic questions about market size, acronyms, etc...? Do those make an appraisal invalid, since you are answering your own question?)

If you feel like debating, PLEASE KEEP ALL DISCUSSION PROFESSIONAL AND MATURE.

Thanks,
-Matt
(Looking forward to reading your opinions!)

 

[JMJ]

A person can't appraise a name and no matter how many algorithms a person puts into a script it could never determine where a persons selling point is or how fat of a wallet a buyer has. There is no way to appraise any name.. Period.. An example would be a recent name I made an inquiry on. The owner told me he recently turned down a $300,000 offer. While this is a 5 letter .com it's not actually even a real word and it gets soso type-ins. He did say he would sell at "a retirement selling price" but otherwise he will continue with his plans. This particular name, if it came before our panel of experts might be lucky to get a $x,xxx appraisal. What makes it worth $300k+? He has plenty of money, doesn't need to sell, and the offerer had plenty of money to spend on it. Long story short and said time and time again. A name is only worth what a buyer is willing to pay and the seller is willing to sell for.

 

[Reppy]

A name is only worth a buyer is willing to pay and the seller is willing to sell for.

At the end of the day that's what it's all about.

 

[user-7256]

A person's opinions can give the owner a general idea of its worth. Some names are very difficult to know. For example, if you own something like ChurchShoes.com, you know it's a good name ... but - you aren't exactly sure of what to ask for it. IMO, this is worth low-mid $xxx without any development or traffic. Some others are more difficult to appraise, especially if TM issues are involved.

The more appraisals, generally, the more accurate your idea of the name is...

 

[Keynes]

Computers cannot possibly know the current market demand

Actually, they can in theory but this is quite an interesting question. Econometric estimates of demand functions for markets are widely used, and can be written in a functional form friendly to computation. There hasn't been any formal work on this topic in the domain market, but a proper specific function could help improve base level value estimates considerably. A demand function w/ a stochastic error term, even if properly specified, however,cannot fully capture all aspects of market demand (since there are certain to be measurements errors, intervals in discrete time would have to be quite small to capture trends/shifts and outliers are quite common in the market given the strong inelasticity of a given name.) Still, I think in working toward base model of valuation, a computed result could provide one of many tools that are useful. We should not rely on models to judge, but rather have them inform a larger set of variables which help determine baseline values -- it's a worthwhile project in either case.

 

[JMJ]

ChurchShoes.com is a great example. You say $xxx. I say $x but wouldn't even register it myself.

 

[user-7256]

(About market demand) ... Actually, they can in theory but this is quite an interesting question.

It is a great concept - one simple way of analyzing market demand is the monthly-updated OVT results... but live results could be much trickier.

Domains as stocks is invalid, IMO, cause they aren't businesses....

 

[barefoottech]

While Computers can give a good approx. value. only Humans buy the names.So various personal choices and prejudicial factors come into play affecting the final Value.
Evidence of this is the usual things you see all the time here and at other forums
(actual TLD's altered to protect people's prejudices)
1/ .@#$ is best
2/ %^#^ is no good
3/ ((((( is only trash

Until these types of Factors can be added and allowed for in automated appraisals , they will not be any where near perfect.

 

[mole]

A name is only worth what a buyer is willing to pay and the seller is willing to sell for.

Ah, the wind that launched a thousand dreams.

Sellers expect more, buyers expect less. Millions of names for resale will end up being retired expired as sellers hope for more, but no buyer will offer more.

Only about 0.0001 percent of all .COMs will end up being sold for $1k or more, imho.

 

[Keynes]

It is a great concept - one simple way of analyzing market demand is the monthly-updated OVT results... but live results could be much trickier.

Domains as stocks is invalid, IMO, cause they aren't businesses....

Actually the models I've been working with have a large number of choice (independent) variables, and OVT is but one of many. There are a large number of other factors included in the model, but sale price is the dependent variable. The model doesn't rely on domains as stocks (equities) at all -- rather it views them as scarce commodities with limited substitutability.

 

[user-7256]

Actually I think MOST, or at least some that I've used take the extention into account and deduct appropriately.

Now what about this - how much data do you have to give the computer before the appraisal is redundant?

I think it's called... Swift Appraisals, where it asks you everything and doesn't do any research itself... it asks you about the market, if there's acronyms, etc. You basically answer your own question and tells you what you think it's worth then turns out to be wrong.

How much should a computer be told?

 

[user-7256]

So... anybody care to discuss this?

 

[gazzip]

I rekon automated appraisals are a total waste of time personally, if you asked 1000 people here what the value of **^%$"!.com was the price would range from reg fee to $XXX,XXX - Who is to say who is correct ?

The price someone is willing to pay is what it is worth ! (to an end user) and you can only guess or hope your domain is worth as much as "similar" previous sales.

JMO - Computers are only as good as the people who design the software

Put shit in - get shit out !! :hehe: way to many variables !

 

[wow!]

Its the brand and the image of whos appraising ....
All posts here so far are against it but why you guys getting GOOGLE PR results as a Bible??? or OVERTURE results as God...???
Thats branding and popularity of whos doing the business.
So I bet when a big firm such YAHOO would make an automated appraisals, you'll change your mind....

 

[wasistdas]

I see no way for automated appraisal to give actual value to a domain name. Too much of human and business factors. And also very dependant on impulse buy.

 

[user-7256]

Hmm... but what about type-in traffic then? You can't ignore OVT results for exact results with a .com - likely it will get almost that much typein traffic - ESPECIALLY w/ extention.

Indeed all your points are valid - so are we saying that all appraisals are worthless? What about professional ones?

 

[JMJ]

Who are the professionals appraising your names? Do you see a list of certified domain appraisers on any site or anywhere for that matter? Most likely they are hourly employees using some sort of automated script. It looks nice to have have a pretty little certificate with a name like sedo attached but as pretty as it is it's worthless.

I've tried and tested the waters as many here have a million times. I get quite amused having a name appraised for $xx by the professionals of the industry here and other places while I'm in the middle of $xxxx negotiations.

Lets use recent activity as an example.

WholeFoods.us $10,000. Now if this name came before you on the boards to be appraised what would you have given it?

PartnerCash.com $110,000 if this name came before you on the boards to be appraised what would you have given it?

The reason people pay these prices is simple. They have a plan, and more money than they know what to do with... A computer will never be able to determine the size of a wallet. It will never be able to determine a persons breaking/selling point.

Lets use another example with the "same names same buyer".
OrganicFoods.us $10k
GourmetFoods.us $10,750
WholeFoods.us $10k

I own ethnicfoods.info
Ideally if you used several factors my name would be worth more than the above. Why? .info is recognized around the world. More registrations etc. Sure it has a bad rep at the moment but only on these boards.

Will I ever see $10,000+? I sure hope so but most likely nowhere close.. Will the owner of the above names ever be able to sell his purchases for $10k+? Most likely not. Do these names make thier company more valuable? No, atleast not indivdually but as a whole mabey?? In reality they're just an asset of an existing company an until the investment makes a return they are a loss.

 

[gazzip]

Actually I think MOST, or at least some that I've used take the extention into account and deduct appropriately.

Now what about this - how much data do you have to give the computer before the appraisal is redundant?

I think it's called... Swift Appraisals, where it asks you everything and doesn't do any research itself... it asks you about the market, if there's acronyms, etc. You basically answer your own question and tells you what you think it's worth then turns out to be wrong.

How much should a computer be told?

You should tell it as much as you can - but what you will never be able to tell it is - what value an individual enduser/buyer puts on it and this IMO is the most crucial one by far !

As stated in above post by NameMogul

Good thread this - there's nothing like a good MASS DEBATE :hehe:
.

 

[JMJ]

What I find curious is how the shizzle does someone even have the guts to hold out for $10K for those .us domains? Where did the buyer start? Did he start far too high? Sometimes appraised or not, stupidity prevail via poor buying skills.

I don't think it was a case of holding out. I believe it was a case of an unexperienced person making an offer. A very large offer. Once the first sale went down that set the standard for that particular buyer. Those who aren't "in the loop" are the ones getting less money.

If not the above then another assumption would be the topic we are discussing. An "appraisal" was ordered either by the buyer or they are using sedo's broker service and buying at the "professionals opinion price." I haven't used their broker service but from my understanding is they set a "value" and you have to make an offer within that range. How valuable would you think a name was worth if someone paid you to tell them and you got paid 10% of the sale in addition?

 

[user-7256]

Very, VERY good reads Thanks for your thoughts! Any more have any opinions to add?