Ask three domain name investors for a valuation of a domain name you are selling or thinking of buying, and you are likely to get three different answers – maybe even four.
Enter Dennis Lastochkin, a technical genius who in just six months created a sophisticated valuation engine that is likely to remove most of the subjectivity from domain name appraisals.
In this interview, Lastochkin discusses:
Dennis Lastochkin is the chief technology officer of DomainParking.com and DomainWorth.com.
Lastochkin is a competitive cyclist and resident of Austin, Texas.
. If you have a great domain name and nothing to show when people visit, you’re missing out on potential advertising revenue, leads, and partnership opportunities. NicheWebsites.com can build you a site quickly with a price option to suit any need — but as their tagline says, they don’t just build websites, they build businesses.
. Value domain names you want to buy or sell, evaluate drop lists quickly and efficiently, and use the lead generator tool to contact potential buyers using saved templates and an automated email process. It will increase your accuracy and productivity by magnitudes.
. DNX.com is a domain name exchange that uses a reverse auction platform to provide fair market prices for quality domain names that are manually filtered by an experienced broker. At DNX.com, domain name prices drop until someone decides the price is right; but don't wait too long or a domain you love might be purchased by someone else.
Watch the full video at:
Three messages before today's interview educates and motivates you.
First, if you're a domain name investor, don't you have unique legal needs that require domain name technical know-how and industry experience? That's why you need David Weslow of Wiley Rein. Go search for David Weslow on DomainSherpa, watch his interview and you can see for yourself that he can clearly explain issues, can help you with buy/sell agreements, deal with website content issues and UDRP actions, and even help you write your website terms and conditions. David Weslow is the lawyer to call for Internet legal issues. See for yourself at NewMediaIP.com.
Second, managing multiple domain name marketplace and auction site accounts is a pain. Inevitably, you forget to sign into one and lose a great domain…or worse. Now imagine using a single, simple-to-use and comprehensive control panel to manage all your accounts. That's Protrada. You can setup search filters, analyze domains, automate bidding, list domains for sale, and buy domains across all major marketplaces. Protrada also has a new semantic engine that builds Google-friendly websites with rich content and network feeds. Sign up at Protrada.com to get 20 free credits and start building and monetizing your domains today.
Finally, if you have questions about domain names, where should you go to ask them? The answer is DNForum.com. Not only is DN Forum the largest domain name forum in the world, but it's the best. You can learn about domain names and the industry, buy and sell domain names, talk about domain name news, and meet other domainers just like yourself. Register for a free DN Forum account and begin advancing your skills and knowledge today. And when you do signup, send me a friend request so we can connect.
Here's your program.
Michael Cyger: Hey everyone. My name is Michael Cyger, and I'm the publisher of DomainSherpa.com - the website where you come to learn how to become a successful domain name entrepreneur directly from the experts.
Everyone knows that property valuation is difficult. Talk to three real estate agents and ask them to set a value on your house, and you'll get three completely different answers. Most likely, one of them will be much higher or lower than the other two; and also likely, there will be no overlap between the three valuations.
The same is true for domain name valuations. Ask three domain investors that you know for a valuation on a domain name that you're thinking about selling, that you're thinking about buying, and you're likely going to get three different answers. We've published an article about online valuation tools like Estibot.com and Valuate.com in the past. And in the comments, not a lot of people were happy with the result of these two tools. I've interviewed Ryan Colby in the past and he detailed fifteen factors for determining a domain name valuation.
Today, we're going to learn about a new online tool that goes beyond the other online tools and we're going to find out how it does it. Joining us to tell us more about domain name valuations and the DomainWorth.com valuation tool is the Chief Technology Officer of DomainParking.com and the Chief Architect of DomainWorth.com web app.
Dennis Lastochkin: Hey there.
Michael: Hey, Dennis. So I actually missed my intro there, Dennis. Your name is Dennis Lastochkin.
Dennis: Lastochkin. That's right.
Michael: Lastochkin. I was so close, Dennis. Dennis, welcome to the show. Thanks for joining us.
Dennis: Thank you for having me.
Michael: Dennis, a lot of people are weary of online valuation tools because they don't know if they can trust the results and they don't know what is in that black box. You put in a domain name. It spits out a value. You promoted DomainWorth.com at the Domain Fest Conference in Santa Monica in February 2012. I came to learn about it because it was actually on my table. You handed these out on the tables.
Dennis: Right. Those were the breakfast.
Michael: Yeah, the breakfast. Exactly. And I showed up. It's like a certificate and it says, 'Domain Worth'. I love the name. Like, here's the worth of this domain. And all of them were different as far as I can tell. Like this one says, 'NYEvents.com' and then it's got a little market valuation high and a price. And another one was Milligram.com. I'm not sure if these are domain names that you own, or they're just ones that you ran.
Dennis: Those were the domain names that were actually at auction. And we kind of tried to have people look at the valuations before they auction and kind of compare the auction results to what we had to say.
Michael: Alright. So you ran valuations during the Domain Fest Auction on the actual sales. How did your valuations, using DomainWorth.com, compare to the domains that actually sold and what those values were?
Dennis: Right. So, the last day of Domain Fest in February, we were actually sitting in the room and while the auction was going on, we had the computer with us and we were checking the valuations. And I can say that about ninety percent of the time, we were in thirty percent margin of error. So, we were pretty good for domain appraisal. It might not be very good for house appraisal, but very, very decent for domain appraisals. And actually some of the domain names that didn't really correspond to our appraisal, didn't sell. So it was like sitting there and seeing okay, is it going to sell? Is it not going to sell? How close is it going to get to the appraisal? And some of the appraisals we had high and the domains started really low and then we're like, 'Oh, man. What's going on here?' And then, woah, the price climbed to about our appraisal price. So it was really interesting to see. And some of the people liked to look over our shoulders and look at our screens and see what happens, and some of them were very impressed.
Michael: Awesome. So basically, within a thirty percent margin. So let's say that I was selling a million dollar house and I used your valuation tool if it worked for real estate. It would have a valuation between seven hundred thousand and 1.3 million if the true value, let's say, was a million dollars.
Dennis: Right. And some of them came much closer than that. I mean, we had a lot of people. We had a booth set up at Domain Fest and a lot of people came over and just tried out our service and appraise their domain names over there. And some of the people said a hundred thousand and it came up as hundred thousand and one or something like that. So sometimes it was really close.
Michael: Well, I've actually run it through some of the domain names that I've looked at, that I own, and I think it's remarkably close as well. It gives you a pretty consistent range. Not just a value, but a range. A low value and a high value. And I appreciate that. Of the real property owners that I speak to, if I talk to them about a house, I don't want to know an exact number. I generally want to know a range. I want you to look at the comps, I want you to make some comparisons to the house I'm looking at or the house I'm looking at selling, and I want to generally know what the range is. And so, I appreciate that about your system.
Dennis: Right. And some people, when they sell a domain name, it depends how motivate they are. If they want to sell a domain name tomorrow, if you want to sell a house tomorrow, you're probably going to take a cut compared to if you're willing to wait ten years. So that range is actually good. It depends on how motivated you are, so you cannot just put a number. You have to have some range depending on how motivated you are and so forth.
Michael: Awesome. So we're going to dig in and figure out how the system works and what you look at in order to come up with this valuation range. But I want to take a step back and find out a little bit more about you, Dennis. Because I'm not sure a lot of people in the industry know about your background. How did you get the skills to be able to architect a system like DomainParking.com or DomainWorth.com? What's your background?
Dennis: Well, I started pretty young. When I was fourteen, I lived in Israel and I took a C++ class. When I fourteen years old, I was kind of biking to that class. At the time I didn't drive a car and you couldn't drive a car at fourteen in Israel. Ao I biked to C++ evening class in (Unclear 6:56.7), and that's how I started. And then I did some homework for my mother, and at about age sixteen, I was already writing small programs. So it was an early start. I was actually born in Moscow, Russia. So I lived in a couple countries before I came to Austin, Texas. As far as how I started with it; when I was eighteen years old, I came to Austin and I met an FM in Eagle. And I'm sure you guys know, Eagle is Mrs. Jello. And I started working for them and we started from a service called IBoogie.com. And it's still up and basically what it is, is a clustering agent. And it's useful to group snippets or paragraphs of text into meaningful erratical groups. So, let's say you have a thousand answers from a survey and you want to know more or less what those answers are talking about. You can put those into our engine and get main topics of what those answers are talking about.
Michael: Interesting. So the URL is IBoogie.com. And if I do an open ended survey, like somebody comes to Domain Sherpa and I have a little popup on the bottom that says, 'Tell us how we can get better' and it's an open ended response. And I don't want to have to go through every single response to figure out what the general trends are and count up who's saying the site needs to faster, and the site needs more technical interviews, and this, and that. I can put it into IBoogie.com and you'll semantically figure out.
Dennis: We have an API that you can use and basically you can use IBoogie.com. IBoogie.com is basically a little more different. It's for search engines. So basically it's a clustering search engine. So you have, let's say, a thousand results or a hundred results and you want to know more or less what those results from a search engine are talking about. So IBoogie.com does that for you. It takes those results - the snippets of text - and clusters them into groups so you can see immediately what the subject is, what the different aspects of the results are, and so forth.
Michael: Very cool. So you worked on the technology for that. The public facing side is IBoogie.com, but then you probably have an API that I can connect into and buy that service or buy access to the service.
Michael: Very cool. So where did you go after working on IBoogie.com? What did you move to next?
Dennis: Well, we moved to actually building a parking system. We're one of the first people that built a parking system. It was called Domain Spa. Now it's turned into DomainParking.com.
Michael: It was DomainSpa.com?
Michael: And it became DomainParking.com, which is a fantastic URL.
Dennis: It's a good domain name. Yes. So, we built a parking system. And after parking system, we moved to registrar, so we have our own registrar. And after that, we moved to an image content management system and drag and drop automatic website builder. And now we've moved to an appraisal. About a year ago we started working on the appraisal and, yeah, that's where we're at right now.
Michael: Wow. So you had your hands in - you worked on - all of those different technologies?
Michael: That's fantastic. Wow. The experience that you've got, Dennis. That's amazing. And at the beginning of the show I said you were the Chief Architect of DomainParking.com and DomainWorth.com. Were you also Chief Architect or were you learning more about architecture when you were working on the content creator and the registrar?
Dennis: Oh, no. I was the Chief Architect to the content creating system and the registrar as well.
Michael: Wow. Alright. So a ton of background. You understand how, generally, the Internet works, how domain names are handled from cradle to grave, you understand the technology that supports all those, so no holds barred. Any type of technical questions that I have, I'm going to throw them out at you, Dennis. Because you clearly know a heck of a lot more than I do and most people watching this show. Alright. So, now that we know your background, let me ask you a little bit about how long it took to build DomainWorth.com. If I go to DomainWorth.com and I type in a few domain names, it spins some wheels sometimes and it shows that you're checking information, and then it gives me a full appraisal online - sort of like this printout that I have right here in front of me. How long did it take you from cradle to grave to build that DomainWorth.com Appraisal System?
Dennis: Well, it's a work in progress. But we started last summer. And I believe we're still just beginning, but it's definitely showing some results. So we talked about the idea last summer and we thought that there are a couple solutions online and none of them suits your needs. And the appraisals that other companies are doing - the manual appraisals - are little bit too expensive and we cannot do thousands of them. Well, we thought, 'What if we come up with a system that does it automatically? Wouldn't that be great?' And we started last summer and we launched it at Domain Fest together with the parking system. Right now it's in BETA and hopefully we get some response and continue from here.
Michael: Wow. So, Unbelievable. In six months you took it from a concept to a fully functioning product, which includes coming up with the specification around the features that you're going to have; the factors that you're going to look at; the algorithm that actually takes the factors and manipulates them - sort of the Y=Fx. You've got factors that go in that produces the Y value - if people understand how inputs relate to outputs - then you get your output. But you also get this range, so you're doing some sort of sensitivity analysis. And then, it's a beautiful form factor. You spit out these nice certificates. You also create a link. So if I'm buying a domain name and I think your valuation is high, I can send you a link to DomainWorth.com that shows you the valuation that I've run on the domain. All of that; six months roughly.
Dennis: Yeah, about that. Maybe a little longer.
Michael: Do you have a team of like fifty programmers and designers? How do you do that?
Dennis: No. We're actually a team of two people that started working on it. Actually me and Christoph. He has a Mathematics PhD. Very smart guy. And, yeah, we started in the summer and obviously we had some background, so we didn't start from zero. We knew it wasn't going to be easy. I remember seeing your show and I think you interviewed this CEO, and he said that back in the day he actually did teases on domain valuation and he came to conclusion that these teases weren't really working properly or something to that point. And we figured, well, some people tried it before and it's definitely not easy, so how do we make sure that we come out on top and we actually have a system that's working really well?
Michael: Yeah. So, you and one other person who's a PhD in Mathematics. I can understand. You actually coded algorithms, not your associate?
Dennis: We did both. We worked with project (Unclear 15:30.2).
Michael: So you're both programmers?
Michael: Okay. Who actually did the user design?
Dennis: The logos? Or the layout? Because it's really simple. There wasn't much to user design to it. I mean, right now, we're just trying to concentrate on the actual function and algorithm, so we didn't put much into design.
Michael: You're selling yourself short. It's a nice design. It's clean. It's fully functional. I understand. Did you actually do the design work - the layout of the web page?
Dennis: Yes, I did. We had some of the files and we obviously had a designer do some of the logos or maybe some of the little elements on the page. But, yeah, I pretty much worked on the design.
Michael: Well, that's very impressive, Dennis. I can think of a few projects that I'd like to work with you on going forward.
Michael: Alright. So, six months to build Domain Worth Appraisal System from concept to fully functional. I understand that things like this are always a work in progress. You might tweak the algorithm just like Google tweaks their algorithm as things change in the market. Have you built the system such that it's a learning system over time so that it gets better by itself with you actually not having to tweak factors?
Dennis: You mean like an artificial intelligence program? Well, we definitely have checks and bounds. And obviously, the more time goes on, it collects more data. It gets smarter. It gets a bigger sample of things, so it becomes more accurate. So, in that respect, yes. I mean, as time goes on, it will become more and more accurate. And I hope that whenever people start using it and they see what the domains are worth, I hope that we actually become, I guess, the domain authority. Like you are the domain authority for interviews, hopefully we become the appraisal authority for appraisals and hopefully, if not change the market, but at least people will be able to recognize us.
Michael: Definitely. Okay. So let's talk about the factors that go into your valuation range. I'm looking at the appraisal of NYEvents.com. It has a valuation of five thousand four hundred and eighty-one dollars - at least at the time when this was printed. And, as we just said, new data comes in all the time and this may change over time; just like house values change over time as well. The market low - $4,228. The market high - $16,304. I can see that you split the name in to words. You look at the registration day. You look at the length of the domain name. You have information about how the domain name ranks on Alexa; Mozilla.
Dennis: It's SEO (Unclear 18:34.5), yeah.
Michael: You look at the search engine. Searches per month. The similarity around other domains. The actual comps that you have in the system. So I'm running them through. Are there any other factors that I didn't mention, Dennis?
Dennis: Well, before, I guess, we talk about the factors, I would like to kind of back a little bit and talk about how it compares to something already existing. Like if I could compare it to a real estate appraisal. Because that's how we came up with the model and I like to explain what kind of different ideas we thought about before we build this and before we go to the factors.
Dennis: So, when we thought about an appraisal system, we looked at what was existing. And, basically, like you said, what's existing in the market today is pretty much, you put a domain name and they give you similar factors, and they pretty much spit out a price without any kind of backing up with some data. Because obviously if you just get a price without anything to compare it to, it's hard to trust it, right?
Michael: It's a black box. You don't know whether to trust it or not.
Dennis: Yeah. So when we built the appraisal, we looked at the real world appraisals and we came down to the conclusion that a lot of people compare domain names to real estate market. Like they say the domain names are basically the real estate of the Internet. So we took the real estate as an example and I actually bought a house five years ago and I got a real estate appraisal. And when I got the piece of paper, basically it gave me the price, but it also gave me a list of properties in my area. And basically each property had certain parameters like square footage, and, I guess, location. And each parameter was adjusted to my house and basically the appraisal was based on those properties. Because you cannot compare a house in Austin to a house in New York City. It wouldn't make sense. So we thought the same way about domain names. It doesn't make sense to compare a very short domain name or a very long domain name. Different domain names have other domain names you can compare it to. You cannot compare it to the whole market. It would be crazy to compare one domain to every other domain.
Dennis: So, the challenge was finding analogies between the real estate and domain name appraisals. Because with real estate it's kind of obvious what the parameters are. Like location and square footage are two very big parameters in real estate. And in domain names, it's less obvious. How do you appraise a domain name? There's no real location. Everybody can access a domain from anywhere in the world. It would be different if you can only access a domain from your city, but there's access from anywhere in the world and there's definitely no location and there's no square footage. So we had to find those different parameters.
Dennis: And then we had to think how people domain names - how people think when they buy a domain name. And, obviously, there isn't one formula for every person. It's not the same. People think of the value differently. It's like you said, at the beginning of the program, you had three people that value a domain name and they would value it differently. So we had to kind of go reverse engineering on people. See what people look at when they buy a domain name and how can we integrate it into some sort of a model. And we definitely did a lot of model building, therefore, the Mathematics PhD. We definitely had to build some of that. So that comes down to the parameters. Which parameters are important, and which are less important, and so forth? And the last thing we wanted to do was to make our platform transparent incredible. So, what that means is, when you look at our appraisal, even if we give you the price that you're not happy with or you think it's wrong, you'll be able to see it from the appraisal because we're trying to be transparent. Even if we're wrong, you'll be able to see it and notice it; and compare it other systems where it's like closed box. So, saying that, I would definitely like to explain what kind of parameters go into all that and go forward.
Michael: Yeah. Okay. So, I went over a high level review of the paramaters. Are there other parameters that you take into account that aren't listed on a valuation that DomainWorth.com produces?
Dennis: This is something we show to the people. Obviously, if we had to show all of our parameters, we wouldn't have enough space on the page. And some of the models that basically build those parameters are in the back end and we're not going to show those. But we are showing what we base our appraisals on and hopefully people will be interested to see what those parameters mean for the domain name and how it affects the price.
Michael: Okay. So, clearly, we're not looking for the secret sauce. We're not looking for the actual formula - the model - that you use. But if somebody does a valuation, these are the parameters that you show when you produce the valuation. Those are the major parameters that you use to produce that number - the final value.
Dennis: (Unclear 25:04.4).
Michael: And from your model building, Dennis, do you notice that some of these parameters are much more important to the value than other parameters?
Dennis: Yeah, obviously. Same thing like in real estate. Obviously location. They say location, location, location. Obviously some parameters are more important than others and it actually depends on the domain names and maybe on the extensions. It changes.
Michael: Give me an example. Give me two domain names. Just make them up and give me an example of how one name would have a higher value with certain parameters and another domain name would have a higher value with other parameters.
Dennis: Well, let's say you take a very short domain name and a very long domain name.
Dennis: For example, for a very short domain name, the number of words, for example, it's really their domain name. It doesn't matter how many words it has. I mean, there are only so many words you can make out of a two letter domain name. But, let's say, it's a four letter domain name. It's two words or one word; it doesn't really matter. But if you have a domain name that's thirty characters long and if it consists of, let's say, fifteen words or something, it's going to be a lot cheaper than if it consists of two words. Maybe they're just long words, but it takes thirty letters to spell those words. So something like that would definitely change it. The number of words would be important depending on the domain name length. Something like that.
Michael: Got it. And with a shorter domain name, like AA.com, let's say, which of course is American Airlines. You'll have similar domain names that have sold recently that are two letters that will have a much higher price than thirty character domain names that have sold recently that have probably, if any have sold recently, have a much lower comparison values.
Dennis: Right. So, obviously, domain length is one of the first, very basic, parameters and I think everybody recognizes it. The longer the domain name is, the harder it is to remember and the easier it is to misspell the domain name. And, just like a trademark, the longer the trademark is, the harder it is to remember your trademark. Same thing with domain names. People love short domain names.
Michael: Definitely. Okay. And one of the first items on here, under domain details, is best split. So, for NYEvents.com, for example, the best split was NY as one word and then Events as the second word. We talked briefly last week or a couple of weeks ago when we were setting up this interview, Dennis, and you said that that's an important factor that you really tried to nail with Domain Worth that other systems may not get entirely right.
Dennis: Right. It actually seems like a very easy task to do, right? You have a domain name and you have to figure out what are the words of that domain name.
Dennis: And what does the domain name consists of. And when the person looks at it, he can do this task every easily and he can definitely recognize what the domain name is about. Although we did have some domain names that we weren't sure about. But when the computer does it, a lot of people that tried it before had a huge problem. Because like, for example, NY. NY is not a dictionary word, for example. So how do you know that NY is a word? And we definitely took a lot of time and a lot of effort to make sure that when you put a domain name, it splits it into the correct keywords and ninety-nine percent of the time, and make sure it does it right.
Michael: So how do you do that, Dennis? Clearly, you can get a dictionary and you can include it as a table within your system, and you can do queries against that table to find out if events is a word. But how do you do NY?
Dennis: Well, I won't be able to tell you exactly how we do it. But basically, what we do have is one of the projects that I worked on, which was to crawl the Internet. And we have a search engine also on our backend. And since we have the whole Internet at hand, we can definitely know what we're dealing with. So, we're not just dealing with the English dictionary. We're also dealing with misspells. If somebody misspells a word, we also want to know if that's a word versus just like a regular dictionary.
Michael: What I found interesting also is, I run an separate publishing company other than Domain Sherpa and I was looking at a domain name recently for a tool that's actually a Japanese word. Because a lot of Lean and Six Sigma manufacturing and services are based on a lot of the great work that Toyota Production System has done post WWII; when we went out there and taught the Japanese some of the quality procedures and then they took it to the next level, and Toyota did a fantastic job. And there's a lot of Japanese words that they use to describe the way that operations should run. So I'm talking to a gentleman about this domain name and Estibot gave the domain name - a two word domain name - a five dollar valuation. GoDaddy gave it a valuation between eighty dollars and three hundred and twenty dollars. And Domain Worth gave it a valuation of twelve hundred and sixty-seven dollars, which has a valuation range between a thousand thirty-nine and two thousand three hundred and thirty-six. And I think, clearly, the Estibot valuation is just off. But I think that your valuation, including the range, is right on the money for that type of domain name. I think GoDaddy was a little bit lower. And I think the big issue was that it was a Japanese word that a lot of these systems don't understand. But for some reason your system did understand it.
Dennis: Right. And like I said, we definitely have the background to do something like that. When we worked on our previous project, we built a search engine. And so, we can leverage that and obviously we have a search engine that covers Japanese words. And obviously all domain names are ascii characters in English, but because we have a search engine, we can definitely split and find the correct way to look at those domain names.
Dennis: And another thing that's very important, like you said, Japanese. It's very important when you look at domain names, not only the lengths, but also you want to look at extension. Obviously different extensions are going to sell for different prices. Like .COM is probably the most expensive extension. And maybe you have a .NET extension, which, depending on how you look at it, sometimes can be five or six times cheaper. And it's important where the traffic is coming from. So, Japan is a pretty rich country and the gross domestic product in that country is pretty high. And everybody in Japan pretty much has Internet, so we factor that into our price. If your traffic is coming from India, I mean it's good traffic, but people can spend less in India because the gross domestic product is lower in India. So it should be reflected in the domain names. Depending where the traffic is coming to that domain name, it's going to influence the price. Not just the amount of traffic, but where it's coming from. And we actually show it. I think we're the only people that show where the traffic is coming from. You have a nice little map on the appraisal and you'll be able to see where the traffic is coming from.
Michael: Yeah. And so, most people, when they think of a domain name appraisal, they're thinking that it's just the domain name and not any website that's on the domain name built out. How does your appraisal system work? Is it only for the domain names, so you're looking at traffic only related to domain name not the website and the traffic associated with the website?
Dennis: Right. I mean, when we look at a domain name, we're just trying to valuate the domain name. We're not trying to valuate the website. Because obviously, valuing a website is a whole different thing. If you want to sell your website together with the domain name, there are so many other factors that go into it. What language the website is written in, how much money it generates, how easy you can transition it, and so many other factors. And in some of the domain names, like CNN or Google, they have websites on them obviously, but nobody's going to sell you those domain names. So, for some of those domain names, it would be pointless to appraise them. Because when you're buying something like Instagram.com and it just sold for a billion dollars, the domain name is probably a very small fraction of the sale price when you're actually buying a website that's probably part of a company. And we're not trying to value companies. We're trying to value domain names.
Michael: Okay. Let's say that I have a website that's on a domain name that has a thousand people per day. Does your algorithm actually take into account those thousand people per day? Or are you only looking at the type-in traffics that are going to the domain?
Dennis: No, we take it into account. When you look at a domain name, you can have traffic from different sources. You can have type-in traffic. You can have misspelled traffic, where the domain name is similar to other domain names and maybe it's in a different extension, or maybe it's missing some characters, or maybe it's another additional character, or maybe the characters are switched, or something like that. Or you can just have links to the domain name. Obviously we take all those into account when we look at the traffic ranking of the domain name.
Michael: How do you get the traffic of a domain name?
Dennis: Well, obviously, unless you park with us, we cannot really get the traffic. And even then, we're not going to show people your traffic just because of privacy issues.
Dennis: And obviously, I have no idea how much, for example, DomainSherpa.com gets visits per day. But we try to do an educated guess based on different factors. Like you said, we have Alexa. We have number of links. We build models to kind of model, based on the different factors, how much traffic your website can get. And it doesn't have to be exactly on the money. It has to be a range. And that range definitely influences the price.
Michael: Okay. Are the factors at the top of this valuation, Dennis, more important than the factors at the bottom? Or is it simply laid out to make it look nice?
Dennis: No. There's no such thing. It's basically we start with the most basic things, like domain length and creation date. And then we go into a little more detail. It depends on how much you want to go into it and then you get more details. So, we lay it out that way.
Michael: I see Alexa Rank on here, so you're looking at how well does the domain name rank in terms of all the websites out there. I don't see anything on here about Google Page Rank. Do you take that into account if a domain name has a value of one (Unclear 37:35.8)?
Dennis: Google Page Rank was something Google had on their toolbar and with time, Google Page Rank became less and less reliable. Until not long ago, they just disconnected it altogether. And I wouldn't rely too much on Google Page Rank. If somebody has Google Page Rank, they probably have it from a couple years ago when they collected this data. And I wouldn't rely too heavily on Google Page Rank. There are other factors that you can. First of all, you can now calculate your own Google Page Rank. And, for example, SEOmoz has the main authority, which is very similar to Google Page Rank, but it's for domain names, so it's kind of like a domain rank. But you can kind of calculate the Google Page Rank yourself if you have the links and the information and you can build a link matrix or whatever you need. And as far as Alexa, Alexa gets their rank from their toolbar. When people visit their website and they have Alexa toolbar installed, Alexa records that and I'm sure they also take data from DNS servers around the world and figure out traffic that way. So we don't just base our traffic rank on one factor. We kind of have a lot of factors together and we try to model based on those factors.
Michael: Yeah. Alright. Sometimes I go onto the Google Adwords Keyword Tool and because I have it connected to my AdSense account on Google, I can see the cost per click information on an exact phrase that I'm looking up that might match a domain name. Maybe I'll see a cost per click of a dollar, for example. But then if I actually go to Google.com and do a search on that phrase, I don't see any advertisers buying ads. You're clearly using that same sort of information. I'm not sure where you're getting the information from, but you have the monthly searches exact, you have the cost per click exact for the exact phrase, for example, NY Events. First of all, do you get this information from Google?
Dennis: Well, we get this information from a search engine. I can't tell you exactly where we're getting it from. But when you asked about why you're not getting results in Google, well, sometimes people have a campaign and the money runs out or depending on geographic location. Google displays those results on many different factors; not just because you type a keyword. They look at your IP Address. They look at where you're coming from. They look at what you clicked before, I'm sure. I mean, I'm not Google, but they definitely have a lot of factors that go into displaying a certain ad to you. So even if you're not seeing something, it doesn't mean other people are not seeing it. And they have different factors, like cost per click or competition. Those are very important. And when you look at those, you try to model to what other domain names have and try to find similarities. And you mentioned the search engine popularity of your split phrase. So, one of the things that we said that was very hard to do and other services weren't doing as well is splitting a domain name. And it's very important to split it correctly, because after you split the domain name, you want to find out what's the popularity of that split and how many times people type maybe that exact phrase, how many times they search for it, or how many times people type those keywords as part of a brother search. And those factors are very important for domain valuation. Because, obviously, if somebody's typing, for example, Ontario Fishing a million times a month, probably Ontario Fishing is going to be a good domain name.
Michael: Well, so it has a great search volume. Maybe monthly searches exact or global searches exact. But it wouldn't have, if nobody's buying advertising, the cost per click would be zero.
Dennis: Right. The cost per click would be zero. The competition wouldn't be as high for that phrase.
Dennis: So we also look at the countries where this search is coming from. And you have to look at the trend. We have a trend graph. And for different domain names, you see that the trend graph looks a little different. And maybe you have a certain domain name that just had a huge spike and now you're seeing it's declining so you're probably thinking, well, maybe that spike was just temporary. And so, you have to look into those factors as well. You cannot just look at an average as it is. You have to look at the trend. And that's very important.
Michael: Yeah. Definitely. If I go to Google Trends or Insights for search, I can lookup the trend for certain keyword phrases over time. You do something similar with that as well.
Dennis: Right. We look at the trends and not just Google, but other search engines as well.
Dennis: And then basically we take that into account. And you know what? The timing of when you want to sell a domain name is also very important. Just like a timing when you want to sell your skis or buy your skis. In the beginning of a season, it might be a little more expensive than the end of the season, where it's on sale.
Michael: Exactly. Or like a house. You'll have more luck selling it in Spring, when people are thinking about moving to different areas for children's schools that are starting the Fall versus trying to sell in the dead of Winter, when everything is dead; where it's covered with snow; where you can't see anything. So there are better times for (Unclear 43:46.9).
Dennis: Right. So, not only do we give you an appraisal, we also give you this information about trends. So if you look at an example, Ontario Fishing, you can see the graph going up and down and basically it's high during the Summer and kind of low during the Winter. And if you wanted to sell a domain name, maybe you want to sell at the beginning of a fishing season. So something like that would be very important.
Michael: Excellent. I don't actually remember seeing that graph, so I'm just going to DomainWorth.com right now and I'm going to type OntarioFishing.com. Let's hit appraise on this. And it's spinning right now; pulling the data. Okay. So I can actually see the search trend. It's actually trending down from 2004 to 2012. So you've got an enormous amount of data in here to look at it over the period of eight years. But, exactly like you're saying, it peaks every single Summer, so you're more likely to get a higher valuation for that if you sell in the Summer than in the Winter.
Dennis: Right. And then, obviously we give you an appraisal, but we cannot tell you exactly what to do. You have to look at it and realize, well, maybe I have a better chance selling it in the Summer at the beginning of the season than selling at the end of the season.
Michael: Yeah. Definitely. Okay. That makes perfect sense. So, how does somebody, who uses your system, know if the valuation is accurate? You mentioned earlier, run the valuation. See if it's wrong. See if it's right. How does somebody know if it's wrong or right? What do you think?
Dennis: Well, we have something that's called Similar Domain Names. And basically we have those domain names that we display a small amount of the parameters for those domain names as well. And you can see what kind of similar domain names we found. So if you're going to appraise CNN.com, we're not going to find any similar domain names and we're not trying to value that because something like that wasn't sold before. Something like that is probably a company. And so, we're not trying to tell you CNN.com is worth ten million, or twenty million, or five million. Because nobody knows and we're just trying to tell you what we know. So, if we come up with similar domain names and you think there's not enough similar domain names or we think we can't find anything like that, then we'll tell you that and you'll be able to see it. And therefore, you can check. And obviously each domain name that we display, we adjust the factors. So you have different categories for domain names, which is very important. And you can see what we've come up with and you can see what we based our appraisal on.
Dennis: Obviously, we're improving the system. In the next couple months, we'll probably even go into more detail and display for people even more information about how we base our appraisals on. And if we get the demand for a certain type of display or something like that, we'll definitely look into it.
Dennis: For now we just have this table of Similar Domains, but I believe within the next couple months we're going to have even more information so people can check us.
Michael: Great. So, it seems, to me, like you've got different models in the background depending on the type of domain name. If you've got a two or three word domain name, you try and split it the best you can; you run a certain algorithm against it. If it's a three or four letter domain name, likely, it's an acronym and so, you have maybe a different algorithm against that since word splitting doesn't make any difference.
Dennis: Right. So when you look at appraising in real estate, obviously you cannot have one formula for all the real estate market in the world. Because maybe real estate market in America works differently than it works in Europe and so forth. And even when you look at domain names, just going manually and looking at four letter domain names and trying to group them, I mean, it's a pretty big headache. So we have a system that actually clusters pockets of domain names and we try to figure out a certain model for this pocket and a certain model for that pocket. And it works much better that way because when people think about domain names and they want to buy a domain name, they think, 'Oh, you know, I actually sold a similar domain name in the past' and they think in pockets. They think, okay, this is a certain group of domain names here - maybe two letter domain names - and maybe this is certain domain names in this category here. And they were selling for this much or this demand for domain names that start with a certain letter. Something like that. And it would be crazy to have the same function for all those domain names, so we definitely try to look at the domain name industry as pockets of similar names.
Michael: So clearly the comps - the comparable sales - are very important to the algorithm. If you have a four letter domain name and you have ten comps that are all within a certain range, you can then make a pretty good valuation based on the recent comps. Clearly if a house that's the exact same size, exact same bedroom and bathroom count, and footprint of the property that sells right next door to your house; and they're both the same age, and the same quality finishes, you've got an exact comp if that one's sold unless there's some mitigating circumstance like your septic feel has failed or what have you. And so, comps are important. How do you make sure that your system has the most comps and the most accurate and up to date comps?
Dennis: Well, a lot of systems have a really hard time collecting data. And that's one of the big problems that people face. How do we get all this data? And that was one of the things that, when we started working on it, we thought how do we make sure we have all those data up to date. And when you go and do an appraisal and it's in your name, you see all those icons are spinning and basically we just get fresh up to date data and not something that somebody sold you five years ago and it's completely different now. We're actually going and getting the up to date data that's right now on the Internet.
Michael: So where do you get that? Do you go to certain sites that provide the most recent sales and then you pull them out?
Dennis: Some of the data we get ourselves. For example, one of the parameters that we thought about would be useful was around comp domain names. So, for example, again I'll go to the example of Ontario Fishing. If you have a domain name - OntarioFishing.com -, you want to see how many domain names are registered around Ontario Fishing. We'll pretend our word is Ontario Fishing. And obviously you would use it with similar names and we found that the more domain names that contain the keywords Ontario Fishing. If you have a domain name that contains more of the split keywords, it will value more. Basically people are trying to register OntarioFishing.com and it's taken, so they register OntarioFishingOnline.com.
Michael: Or MyOntarioFishing, or AOntarioFishing, or TheOntarioFishing.
Dennis: TheOntarioFishing and all that stuff. And obviously they all want OntarioFishing, but they can't because it's taken. So when you sell that domain name, you have all those other people that are interested in this name. And it will be a great tool for people looking to sell a domain name and they can sell to all those other guys that have pretty much registered around you.
Michael: Potentially, it shows there's a lot more interest around those keywords.
Dennis: So that's something, for example, we collect ourselves. Some of the other parameters, we obviously go out to the Internet and maybe call the Internet, or look in search engines and so forth.
Michael: Yeah. Okay. So, for me, if I'm running my own search. I want to buy OntarioFishing.com. I'm running a search. I look at your output. The domain details is information that should be indisputable. You're splitting up the name; either it's correct or not. The creation date, you look it up. The traffic ranking. The search engines. That's just data you're pulling. If I go there, I'm going to see the same data most likely unless there's some error that I'm doing or the system is doing. But the real thing for me to determine if it's accurate is to look at the similar domain names. If I know of other domain names that have sold around it that aren't necessarily public information or that I just don't agree are the right sales to be look at, that's where I might say, 'Okay, it's undervalued or it's overvalued'.
Dennis: Right. And you have to consider those domain, as we show, are not what we use exactly for the appraisal. Because obviously there's a category there and obviously we adjust those domain names based on those categories. And obviously you see there's a country there, where the traffic is coming from for those domain names and we adjust based on that. You see what's the gross domestic product from that country. We adjust based on that. And another thing to look at is influencing domain names. So, for example, for OntarioFishing.com, it's a parked page and I actually own the domain name. Well, I figured, I might as well use what I own. No other people are going to come and complain about my examples.
Michael: No, I appreciate real examples.
Dennis: But it's a generic domain name and there's a domain name, OntarioFishing.net, that's up and running. So, obviously OntarioFishing.com is going to get some of the traffic from OntarioFishing.net. It's the same domain name; different extension. I haven't actually seen any appraisal service that does that. That looks at misspells. If your domain is a misspell, we get that. We catch that. If your domain name is getting traffic from other domain names that may be different extensions, or maybe they are spelled different, or have one letter longer or one letter shorter, we take that into consideration for the traffic, and we take into consideration for the appraisal.
Michael: Makes sense.
Dennis: So on OntarioFishing, you see we have OntarioFishing.net and OntarioFishing.org. And we take that into consideration.
Michael: Definitely. Okay. What about all the new TLDs, Dennis? I just had Stuart Lawley on the show, who is the CEO of ICM Registry. They've got .XXX. If I type in DomainName.xxx, will your system appraise it for me?
Dennis: It probably will not now. Just because, I'm not going to lie to you, there's just not a lot of information to appraise a DomainName.xxx. I mean, from what I understand, those guys opened up this year and they had an auction for all the premium domain names. And I don't think there's even after market for .XXX domain names. So, I mean, I can tell you that, yeah, I can appraise those domain names, but I guess we just don't have enough data.
Michael: So, we don't come back with an appraisal, it'll split the words, it'll look up the Alexa Rank, it'll look at the search engine volume and cost, but when it comes to the comps, if there's no comps out there, then the appraisal is not as accurate.
Dennis: Right. And we can extrapolate. And actually we do in some of the other, less popular, extensions. We can pretty much think about what it would be if it was Worth.com and we can look at different extensions, but at the end of the day, if there's no after market in .XXX domain names, I mean, I would hate to lie to you and tell you, okay, it's worth that much where I cannot base it on anything. I can tell you a formula that is my personal opinion, but then it wouldn't be objective.
Michael: Right. Definitely. And that's what I appreciate most about DomainWorth.com. Is the objectivity of it. You're removing a lot of the subjectivity. When I talk to people about the most important things, they say, 'Oh, this and that unless it's this or that' and here you've got a formula; you're running through a lot of different areas. It's a black box still, but you talk about all the importance of it, and you show the trends, and you show the values, and people know where you're getting the data from. What was the biggest challenge to build out the system, Dennis? Or what was the biggest lesson learned during the buildout?
Dennis: What was the biggest challenge? Well, obviously, when we started in August, the biggest challenge is getting the data. How do we get all those data? And it's a huge challenge even for a lot of domainers - just getting all those data. It's something that we worked really hard on and it's something that's not easily done unless you have background in it and built services to get a lot of those data. And another challenge was, obviously, coming up with the solution of how you're going to go about it. We talked to a lot of different people and every one of them has a little different sauce of how they would valuate their domain names. Like you said, all of them have different formula of how they value the domain name. And we had to kind of go into people's head and try to figure out. When you're doing an appraisal for all the domain names, you have to kind of consider everybody and how they're thinking about their domain names. So that was a big challenge; trying to figure out ways that we can take on something that's already an existing technique that's already used and apply it to a wide market of all the domain names.
Michael: Yeah. Definitely. So what are your biggest challenges going forward with continuing to evolved DomainWorth.com and make it more accurate over time?
Dennis: Well, we're still a small company, for example, compared to Sedo. So one of the challenges for the next (Unclear 59:53.4) of the future is marketing. And hopefully this interview will help, so people now realize that we exist and check us out, see if they like us or not, and give us comments. We would like to get some feedback.
Michael: How do you want people to give you that feedback?
Dennis: We're going to have a feedback form on the website and people will be able to send us their feedback.
Michael: Great. So getting the word out. Doing the marketing. Getting people to use the system. Enjoy it. Get use out of it. Provide you feedback. You're really moving into the next phase now where you've spent a lot of time in development, now it's time to actually get out there and put it through it's paces.
Dennis: Right. Hopefully we maybe get some partners or hopefully we can do whatever we need to make sure people know we're out there and people can use us. And another challenge is continuing to evolve the system. Obviously we're not finished here. We're going to have another version coming up in a couple weeks, which is going to be even better. We're constantly testing different methods and looking for improvements. And that's a challenge. Keeping it evolving and making sure we make the appraisals more and more accurate.
Michael: Yeah. What's changing in a couple of weeks? What's the new secret sauce in there? What's the area that you're improving?
Dennis: You just have to see.
Michael: Alright, Dennis. So we talked before the show. DomainWorth.com. I use it all the time right now. I love the functionality. I can paste in like ten domains I'm looking at and it'll run through and it'll give me a summary of the valuations range. And it's all free right now or I'm signed in so maybe that's why I have access to do that. But it's not going to be free forever, right?
Dennis: Right. Like I said before, we worked on the domain parking platform. So we actually built DomainParking.com. And if you're going to park with us, you're going to get all those appraisals for free. And I would probably love to talk a little bit about the domain parking as well. And we're going to come up with the model. For people that just want to appraise one domain name, that will be free. They'll be able to appraise their domain name or something like that. And if they want to appraise more than one domain name and they don't want to park with us, then we're going to come up with a model that they can pay and they will get our service.
Michael: Yeah. Okay. So anybody that's watching this show. You want to do some valuations. Get over there now. Run through the paces. Provide Dennis with some input. If you don't think that something's right on a particular name, or on a particular group of names, let him know about it because that's how he's going to improve this in the future and I'm sure it'll be appreciative of constructive feedback that you may have. So let's talk a little bit about Domain Parking, Dennis. So, you describe earlier that you've worked on Domain Parking. You're the Chief Architect. Chief Technology Office of it. You understand how that works. And I've talked a lot about domain parking. I'm probably going to have to have you on another show just to talk about DomainParking.com. But at a high level, we talked about what you do differently on DomainWorth.com from other valuation tools. What do you do differently on DomainParking.com that other parking companies don't take advantage of or don't do?
Dennis: Well, one thing we have that other parking companies - and we were actually the first ones to start with it - was displaying relevant media to the user that comes to our parking pages. And what I mean by media is, I'm talking about pictures. So when the person comes to a domain parking page and he just sees results or related phrases that he can click on, that's nice, but after a while, all the parking pages look the same and people recognize that immediately. They see the (Unclear 1:04:08.1). They see those results and it kind of looks boring, so it lowers the RPM for that parking page. What we have different is, like I said earlier, we built content management system for pictures. And we built a search on top of that. So, when you park with us, you can type keywords that you think your domain name is describing and obviously a lot of the work we do ourselves. Sometimes you don't even have to do anything. We'll just figure out the keywords ourselves. And it gets relevant images to your domain name. And it's not just a template. We actually have a huge database of images. And for OntarioFishing, we're going to have exactly the right images for fishing. Or, for example, if it's certain show or certain city, it's not just going to be a generic picture of the world or something like that. We're actually going to have pictures related to that city, or related maybe to that show, or that game, or whatever it is. And that converts a lot better.
Michael: Yeah. I just typed in OntarioFishing.com, which is the domain name that you own that we're talking about as an example throughout this, which I think is useful. And as soon as I brought it up, it shows a picture of somebody fishing outdoors. Probably up North, because it looks like he's wearing a lot of clothes. And it's colorful and it catches my eye. It looks like a parked page, where there are related searches, but there aren't ads displaying right now. There's Walleye Fishing; Manitoba Fishing; Weekend Getaways in Ontario.
Dennis: Right. And when you're talking about domain parking, the different monetization systems that you can put. You can put either a one click system, two click system, or a zero click system. And right now that's a domain name that's basically and I'm testing it. And you can have different systems working on it. Right now what you're seeing is a two click system. So basically you would have to click on the related keyword and then you get results.
Michael: That's interesting. And then the lower level page has another picture on the bottom. So you're trying to draw people's eyes down to a different area as well.
Dennis: Right. I can give you definitely other examples of pages that do have different systems on them. And obviously different domain names are going to have different pictures and different optimizations that fit them better. If you have a very broad domain name, you might not want to display result immediately because people are looking for different things on those domain names. But if you have a very specific domain name, people will definitely be probably better off with seeing advertisement immediately.
Dennis: I can give you this domain name. For example, you can look at MusicTeachers.net and you can see their results immediately when you go there.
Michael: MusicTeachers.net. And it actually pulls in results based, it looks, on my IP Address. So you know my general area, so it's pulling in K12 After School Programs at K12.com/Washington. So you know I'm in Washington and you're not serving up ads that are in a different geographic area.
Dennis: Right. And you see, on top, we have a link that says, 'This domain name could be for sale'. And we have another system for you that will manage your domain sales. So when you go there, you can make an offer for this domain name and when you park with us, we have a whole system that will manage your offers. You will be able to see where the offers are coming from, if this person made a previous offer, and basically we have a whole sales platform for you together with the domain appraisals.
Michael: Yeah. Nice. And so, somebody can click through on that. It takes them to a landing page of, it looks like, your domain names for sale that you manage through DomainPool.com, which is another site that Traffic Media owns. And then people can look up the appraisal, which then ties directly to DomainWorth, and make an offer or you may have a price set already.
Michael: Sleek system.
Dennis: Yeah. We're trying to kind of come around and have all the services integrated into one system so people don't have to go to different places to get their information.
Michael: Yeah. Excellent. Alright, Dennis, the final question is this. People want to know when you're pushing out new updates. They want to know what's changed. Just like there's a whole slew of people out there - hundred of thousand of people; maybe even millions of people - that follow Google and wait with baited breath when they hear that a new release of Panda is being pushed out to certain services, or this or that. I want to know when you're updating your algorithm. I want to know, in general terms, what's being updated. How can I find out when you're doing that?
Dennis: Well, for DomainWorth.com, with every new release, we'll definitely put a little newsfeed. And same thing goes with DomainParking. And I'm sure whenever we're going to release something, I'll shoot you an e-mail and you'll squint it over.
Michael: Good. And then I'll retweet it. Do you have a Twitter account, Dennis, that you use to update stuff that's going on and if people want to follow you?
Dennis: Me, personally, no, I don't have a Twitter account. I don't think DomainParking.com has a Twitter account, but this is something we'll definitely look into. We definitely have, I think, a Facebook account.
Michael: Alright. Great. So we'll link whichever one you have underneath the interview so that if people want to stay up to date, they can stay up to date with you. If you have a newsletter, we'll get that out as well.
Dennis: I'll shoot you that handle.
Michael: Great. If anybody in the audience has a follow up question for Dennis, please post it in the comment below. We'll ask him to come back and answer as many as he can. This is the point in the show, Dennis, where I urge the audience, as I always do, if you received value out of this interview, if you better understand domain name valuation, if you learned about a new tool that may fit into your tool chest and help you do a better job as a domain name entrepreneur or domain name investor, find a way to say thank you to Dennis. He took time out of his busy schedule. He came on the show. We talked about his program in detail; what some of the limitations are that you may want to keep an eye on; how it's better than certain other tools. Find a way to say thank you. It's as easy as sending a Tweet on Twitter, or posting a comment below on the show, or liking them on Facebook. And every time you do that, you form a relationship. I'm going to say thank you right now to Dennis by mentioning his business websites again. DomainParking.com. DomainWorth.com. Both of which, of course, are a part of TrafficMedia.com and DomainPool.com. And all the other sites that they provide that, that seem like they're fantastic and very well integrated. Dennis Lastochkin, CTO of DomainParking.com and DomainWorth.com. Thank you for being a domain sherpa, sharing the details of how you built DomainWorth.com, and thank you for helping others become more successful domain name entrepreneurs.
Dennis: Thank you, Michael.
Michael: Thank you all for watching. We'll see you next time.
Watch the full video at:http://www.domainsherpa.com/dennis-lastochkin-domainparking-interview/