There seemed to be a lot of excitement in the domain industry a couple of weeks ago when VeriSign Announced it’s Domain Score product and made it available to the public.
One blogger even wrote about it calling it: “VeriSign Releases Domain Traffic Treasure Trove to the Public”.
We had a chance to play with some of this data on a registrar only provided basis in the past and were never though the Score’s were an accurate and consistent reflection of traffic.
But when VeriSign put it out to the public, we thought we would take a second look at this “Treasure Trove” of traffic information.
So we ran all 44,625 free drop domains that were dropping on July 8th, 2011 (two days after the product was released) through the VeriSign Domain Score script and to our amazement 38,593 of those domains, well over 80% came back with a 7 day, 30 day or 90 day Domain Score of 2.5 or more.
There were 37,787 scored 5 or more.
Less than 7,000 domain names came back with a zero score.
Now anyone who has been in the domain game for a while, knows that impossible for 80%+ of all dropping domains to have traffic. Actually most of us know that well over 90% of drop domains have little to any traffic.
So if the script is giving 80%+ of drop domains a score, to me it mean the scores are pretty worthless.
We also found of the 2,500 domain names dropping that day with the highest Domain Score, less than 250 were and are still registered as of the date of publication.
Strike Two for Domain Score.
“one blogger” wrote? You can use my name, Michael 🙂
I’m glad you run some tests on this data. I’ve tried much more limited data tests in the past.
What I suspect is that a lot of the NXD traffic is from bots pinging the domains after expiration. You also have a lot of web site traffic hitting some of these. Most of what I got in my (admittedly limited) tests over a year ago was 404 traffic.
My guess is the scale is logarithmic, which would mean a score of 2.5 is very little traffic.
If you started at a score of 5 rather than 2.5 still 37,787 scored out of 44K
maybe even worse
Nice one, @ MHB!
page howe says
nice work, very helpful post.
A little off subject…
Speaking of “Verisign”…
I thought I would mention this…and say congrats to “MHB” & “Right Of The Dot”
“RIGHT OF THE DOT and Verisign Enter into Marketing and Referral Arrangement for new gTLD Applications”
Thanks we are looking forward to a long and fruitful relationship with Verisign
I think the data is probably completely accurate, Im guessing they extract the traffic volumes from DNS root server requests. This includes everything from bots, pings, a multitude of email spam, scrapers, misdirects, redirects and allsorts of other wizardry going on – and the occasional stray human.
If that is the case then the data is of limited use in assessing the likely human traffic of names, as they are not releasing page view attempts by browsers based on http header requests
I’m Michaela and I’m from Verisign. Thank you for taking the time to try out the tool and to give feedback. While different people have different benchmarks on what’s good, on a scale of 1-10, many have told us they only look at names that score a 7 or above. And, DNS traffic (which DomainScore uses) does not equal web traffic. Our scores work to compensate for machine traffic like bots and search crawlers that nearly all domains are susceptible to. For more detail on how the score tries to do that, see this explanation http://bit.ly/nzcmay. We also recommend DomainScore be only one tool of many to evaluate a domain prior to registration.