I think it works great! I couldn't think of any other suggestions then the ones mentioned already, but I can see using it as a reference guide to appraisals I do myself.
I see it as a calculator that can do the first steps in an appraisal faster, after which you would take traffic, backlinks, etc into account yourself.
__________________ § Want to sell domain names- dnReseller.biz is powered by GoDaddy, but costs 40% less... § Selling: AWZR.com/AWZD.com, gamer.vg, C0.org.uk 6M.org.uk 5P.me.uk (PM me) § Discounts on European and Asian ccTLDs
Thanks for this. I've updated the algorithm - I think NNN.tld domains are now valuated properly across the board
Please try it and see what you think.
Keep'em suggestions coming! There is still much work to do with EstiBot, and I'll do my best to keep improving the system according to your wishes
Cheers
Josh
__________________
Your Search For Available Domains Is Over. Estibot Long Tail and Available Domain Tools.
Enter The Final Frontier of Hand Registration Now. The Train is Leaving. DrugIndex.com ~ Hypnotherapist.com ~ MedicalDoctors.net ~ LaplandTourism.com ~ TandemParachuting.com ~ EcotourismHolidays.com ~ Rollator.org
StockMarket.com - 559,282 (a bit low)
vs
StockBroker.com - 26,316 This needs some tweaking , If someone wants to sell me stockbroker.com for 26k i'll buy it tomorrow.
Money.com - 1,491,437
vs
Business.com - 1,989,859
Sorry but I had to run them out of curiosity, interestingly the second one came out on top.
S*x.com - 2,786,006
vs
P*rn.com - 7,613,775
Domain.com - 6,790,678
vs
Name.com - 6,790,999
I would also say the valuation of one word .in is way off but as I don't own many I can't really call you on that.
For example on the .in
India.in is listed at just - 2,899 dollars. Which would be the buy of someone's life imho.
All in all its a good step up from the other auto appraisals and I thank you for the chance to test it but it still misses various factors out of the equation. I realise some of these are impossible to ever do in an automated appraisal but as your developing what seems to be a professional and well built tool you might want to consider these factors.
Specific Industries! Think a gardening hobby name over a financial banking name. (If you could have a drop down box for this to take it into account that would be a further step up)
Market Price Trends - Critical
Is the domain attached to a developed website - Critical
Adwords Scores - Critical
Searches dropping or rising. - Critical
Uniques per day. - Critical
PPC Its earning. - Critical
Advertisers going for the terms.
Ease to misspell the words.
Can you still obtain the plural, .net, .org and .tld of any relevant country.
Country your selling to, is it a national, international or regional name for instance.
How many names are still available to be registered that can perform the same function?
How much competition does the name have undeveloped, even more importantly in the price range your looking to sell at?
How old is the domain? For SEO purposes.
Does the name make grammatical sense
Is the name brandable?
Is anyone advertising the term on TV or through other mediums?
Was the domain a previously existing website with backlinks or are those links fading over time?
How much development has been done around the terms already?
Hope some of them make it into your project at some stage but either way again thank you for a new, professional tool.
__________________ Please click here to see a selection from our portfolio. Our Public Showcase: SharesTrade.com - TextTelephone.com - SecuritiesAccount.com - GoodsSupply.com
eOverdraft.com - TradeDistributors.com - EuronextTrader.com - ContractTerm.com
Last edited by markmiddleton : 05-29-2007 at 10:42 AM.
Thanks much for your comments Mark- much appreciated, and you make very good points.
Before I go too verbose here may I point you to the About, Guide, and Metrics sections on estibot.com. Many of your points are discussed there in some detail.
EstiBot really is optimized for the lower end of domains, as the instructions also say in the Guide and About section. It is really meant for the low- to mid end range, because I feel that's most useful to most fellow domainers. It can usually say if your name is worth hundreds of thousands, but it does not attempt to reach any great accuracy up there, because it's just not very feasible.
StockBroker.com is undervalued I agree, but this occurs due to the stringent error handling routines, and (this can also be found in the instructions on the site) EstiBot is programmed to generally err on the conservative side if it is not sure whether a domain is really valuable or is getting an erroneous overvaluation. I am looking at solutions to these undervaluations but they do occur. Also, whoever owns stockbroker.com will not need estibot to tell the value.
In general EstiBot is mainly aimed at regular domainers working in the reg fee - to low $xx,xxx range. It is meant to be a useful tool, rather than just a novelty site. At least that is my ambition. It is of course fun to test it with obviously mega valuable names, and often it is still in the right ballpark.
All of the above are actually part of the EstiBot system, which is quite a complex algorithm really. It all starts with evaluating the name against previous known sale prices for similar domain names, and goes from there to semantic, keyword and market analyses. The output that you see only lists few of the factors that EstiBot uses in the valuation.
Of course, it is in beta and much work is to be done yet.
Some of the features you mentioned are not in the system and I'll do my best to add as many as possible! Thank you for your comments
Cheers!
Josh
__________________
Your Search For Available Domains Is Over. Estibot Long Tail and Available Domain Tools.
Enter The Final Frontier of Hand Registration Now. The Train is Leaving. DrugIndex.com ~ Hypnotherapist.com ~ MedicalDoctors.net ~ LaplandTourism.com ~ TandemParachuting.com ~ EcotourismHolidays.com ~ Rollator.org
I have done some further testing with it, It is by FAR, The best automated valuation tool i have ever seen, adding your link to AllDeleted.com ...Josh_1
__________________ QueensrycheThey Are Bad Boys!
"You always pass failure on the way to success
Best I've seen, no doubt. I plugged in Casino.rw (which I just won from OUD) and it returned $1377. I put in Mitigate.net (which is up for auction at Sedo right now) and it returned $472. Both in the ballpark IMHO.
__________________
★★★ Click here to see all my domains for sale at Sedo! ★★★
Plugged in some names I've sold with Sedo to compare:
Scrolls.com ($9k) came back with USD $127,409
HorrorForum.com ($650) came back with USD $3,180
SearchKeywords.com ($1250) came back with USD $46,000
Magnify.net ($1500) came back with USD $491
Thanks for testing, guys - all your comments help me improve it during this public beta.
I fixed the mp3 problem, the keyword analysis no longer separates the numbers from keywords so you get appropriate results for mp3 and such names. Thanks to RCRiver for pointing that out.
Regarding individual domain name prices, the sale prices of individual domain names are, in the end, by their vary nature unpredictable, depends in the end on the seller and buyer. For instance, very good names are going for nearly reg fee in the Bazaar. On the other hand, some domain names are being bought for nearly ridiculous prices by some of the big guys.
Therefore EstiBot is, and has to be, a ballpark Estimate, and as such it's working pretty well, although of course it does get it wrong from time to time due to the complexity of the task at hand.
EstiBot tried hard not to inflate or overestimate prices, although it is programmed to estimate the potential value and therefore often gives a higher price than the going reseller value. It is more "sedo" or "afternic" prices that it tries to estimate, and not "bazaar" or "forum" prices.
EstiBot 1) assumes that a sale takes place, 2) puts itself in the position of a motivated buyer and gives the value that it would be willing to pay for that domain name - that sort of summarizes the philosophy.
The valuation is meant as a rough ballpark suggestion, a starting point for critical human evaluation.
A very nice domain name, you definitely sold it at the reseller price. The EstiBot valuation seems a bit high, but of course it's just an indicator that this is potentially a very valuable name.
Good appraisal. This one could have gone either way I guess, depending on the motivation of buyer and seller. $650 is not a bad price, either.
Definitely a nice name and not a bad sale, EstiBot appraisal maybe somewhat optimistic, but this domain name has very, very strong keywords.
Nicely in the ballpark. Good sale. The .net tld is quite unpredictable and difficult to program - some .nets go for nearly .com prices, and others at much less.
Thanks for sharing these - everything helps me to make it better
Josh
__________________
Your Search For Available Domains Is Over. Estibot Long Tail and Available Domain Tools.
Enter The Final Frontier of Hand Registration Now. The Train is Leaving. DrugIndex.com ~ Hypnotherapist.com ~ MedicalDoctors.net ~ LaplandTourism.com ~ TandemParachuting.com ~ EcotourismHolidays.com ~ Rollator.org
Josh-
Many many thanks - I believe this is a paradigm changing project. I have run through many names in this last day - am pleased to find a lot of value that i was not expecting. Perhaps the LLL.mobis are a bit undervalued, particularly good letter combos. I felt the site moving under my feet, changing as I used it -- I thought you said you were taking some time off???
One issue I have found is that piles of junk results on Google will push an ordinary domain way up. Coded PDF files and micro Biology experiment results both seem to generate short letter strings that are used repeatedly on many websites, filling up Google. But Nobody will pay $500 for the name of an obscure newly discovered protein.
Anyway my dream list would include a new search box below the results to save my back button, as mentioned. I wonder if there is a way to identify for the searcher the over-riding reasons for a domain's value. You give title, anchor, etc but something like "heavily searched geographical name in high value region, goes well with .info extension" --- something like that, what ever the top factors your program used for that domain. It is sometimes very hard to understand why values are so different between names. But First:::
Thanks for the quick and detailed response to my post. I have since run a few more of my domains through to test them for you, here quite honestly are the results:
IndexProduct.com, a financial term and often used retail term, came out around 3k and things like SecuritiesAccount.com around 1.7k. Personally I think they are the wrong way around because of the practicality and potential of the names above anything. (the potential of PPC from banks advertising campaigns as well, which exists on both but the account is much more likely)
All these were much more accurate and down to earth approximations than other Appraisal services have been, either rating my names wildly or just giving me conflicting results that seemed very illogical. I hate to say it but so far you've blown the other widely used auto appraisal service, for accuracy on mid range names, out of the water imho; as I only agreed with about 30% of its results, compared to about 60% or more of yours and your only in Beta too.
I know the market has gone a little quiet on e and i domains but I think you need to factor the i and e prefix domains into the equation somehow:
iTunes.com - 4,193
eFinance.com - 4,775
iFinance.com - 1,771 (Extremely low)
eBroker.com - 2,429
iTrader.com - 895 (I think as a general rule trading, trade, trader names need a small boost)
eBank.com - 4,216
etc
Still I think you were closer with my eS*xyVideo.com appraisal about 100 dollars. So maybe its just higher range names again this is happening on.
I think prefix (and abbreviated) names in general are a little undervalued for what they represent but i'm not sure how you'd fix that, unless you had an option for someone to state that at the beginning and what it was an abbreviation of perhaps? I have a few unreleased strong themes I could run right now and show you clearly but can't release publicly for one reason or another, perhaps I can PM you in a few weeks with a series of names that will demonstrate my point more clearly if it will help.
Just trying to help, seems like your doing a great job. If you'll permit the suggestion, here's what i'd do to increase your revenue for your effort:
If someone checks a name, have any extensions not registered come up nearby the result. With of course your affiliate/reseller link so they can register it and earn you a commission. (Perhaps even an affiliate for sale link to the relevant auction site)
__________________ Please click here to see a selection from our portfolio. Our Public Showcase: SharesTrade.com - TextTelephone.com - SecuritiesAccount.com - GoodsSupply.com
eOverdraft.com - TradeDistributors.com - EuronextTrader.com - ContractTerm.com
Last edited by markmiddleton : 05-31-2007 at 07:39 AM.
Thanks again Mark. You are quite right about the iDomain and eDomain undervaluation, I have been looking for a solution to valuate these correctly and will keep doing so until I find one. Will let you know when I have it. It's difficult because as you say iBanking.com is far more valuable than iToiletry.com and that needs to be factored into the algo somehow.
As for the market for these, EstiBot is market sensitive and while I completely agree that these great finance related names you mentioned were valuated very conservatively, the valuations may reflect the market situation somewhat.
By the way, EstiBot valuates itunes.com at USD 323,068 when spelled without capitalization. Of course itunes is still worth more, but according to EstiBot rules any valuation that high is only a pointer in the right direction and whoever owns such an immensely valuable name will need to seek other means of finding out the real potential, whether much more or much less.
It's the downside frankly of my caps system that users sometimes need to check both capitalization styles to find the higher value, but on the other hand this system enables foreign language domains and is far more valid with short brandables and 1 to 4char domains than auto-keyword detection. I am looking at adding a feature that would automatically check both ways and report to the user both values to save the user's time.
Re indexproduct and securitiesaccount, I would advice treating an 1.7k and 3k valuations as similar - you cannot have an automatic system that would reliably make a difference between a domain that's worth 2k and another one that's worth 3k. I would argue that no third-party human can either - it's up to the seller and buyer and their motives. I think indexproduct is slightly overvalued because the wording is not in natural order. I have already partially solved this "natural language" problem (ProductIndex.com gets USD 8,723), but I'm still working on perfecting it.
1.8k for modernfashion.com seems somewhat conservative. A beaut of a domain name. EstiBot taking the role of reseller I'm sure Don't sell too cheap!
I am currently doing an analysis based on the beta test phase so far, and I will be honing the algorithm accordingly.
Most reported problems have had to do with EstiBot underestimating really good domain names. Due to the limitations of the tools that EstiBot uses, the algorithm at this point has to go one way or the other - too careful or too lax, and I prefer the former. Of course, some names slip through the safety net and get a too-optimistic valuation. Working on it, working on it....
Please remember that if you do own some obviously great domain, don't worry if EstiBot gives you a too-low appraisal, it's just unsure of how to cope with the results and usually takes a very conservative position in such a case.
I'd like to thank everyone for taking part in the beta test, this will really help me to make this tool better.
Thank you again for your valuable observations and suggestions!
Cheers
Josh
__________________
Your Search For Available Domains Is Over. Estibot Long Tail and Available Domain Tools.
Enter The Final Frontier of Hand Registration Now. The Train is Leaving. DrugIndex.com ~ Hypnotherapist.com ~ MedicalDoctors.net ~ LaplandTourism.com ~ TandemParachuting.com ~ EcotourismHolidays.com ~ Rollator.org
