Interview with Michael Tanne, CEO of wink

Wink is a social search engine. Its a search engine that finds the most relevant and interesting results determined by what people are doing. We apply a technology, which we call PeopleRank to analyze input from users and determine basically the hand picked searched results based on user feedbacks. So people tag or bookmark sites and web pages that they like and those then we help others come along who want to find those sites. Additionally because we?ve got people involved in the equation we are able to block spam so that the search results are more spam free and also it’s current because it’s changing all the times as people discover new things.

I have been in the search field for a long time and I always wanted to see what would happen if you could have users participate in the search equation. More recently when we decided to actually start Wink I had noticed that you know search results have a fair bit of spam, this is ongoing arms race between the search engines and the search engine optimizers who want to try and get their results to the top. And you know sometimes users are the ones who you know have to kind of put up with that. I mean there are several great search engines in the market and they certainly do a great job but then you know there is got to be innovation and we got to have continued to you know advance of the technologies. And right now in the web what we have seen a lot of is user contribution. So you have got many different examples from Wikis to blogging to social networks and other things where people are participating and so we though well, I wonder if we can take that energy of people wanting to participate in the web and combine that with technology to do something interesting in search.

Absolutely.

Well we are well on our way. What did we do is we have actually noticed that there was quite a bit of user interaction in the form of social book marking already going on. And so what we have done is to call or access a lot of the social book marking services as one way to kind of gather up the websites that people find interesting then they return of course we pass traffic through to those sites these people are interested in the results on those sites. So we actually include the results from a lot of those services and that’s kind of, it helped to get us kick started and then we have our own users obviously having a lot of interaction with the website and so it I think you know where it?s heading towards critical mass fairly quickly.

Well actually you have actually stated it is exactly right that is the crux of the question. You know the other search engines have algorithms that are quite successful. I mean the service that we provide includes both the user input and the algorithms. So we actually do partner with Google. On the Wink side you will see Google results augmented by results from our user base. And so we think that Google is a great platform on which to work and then we are sort of augmenting it with this additional information.

I think that it does work, but it isn’t where it needs to be. It?s a conceptual pursuit and I have been watching relevance algorithms and technologies that approaches this type of search for 13 years. And you know you start with you know the algorithms of the pages and the docs and you know the words and word combinations that are in documents you know Google pioneered the analysis of the link structure between pages. People have explored clustering and natural language parsing and semantic analysis you know Ask has the broad narrow concepts. And you know, and we have introduced the social search concepts of people rank. So I think every one should be always trying to innovate.

I think yeah every successful service uses some blend of those of all those different techniques. And you know it?s usually in terrible iterations as they evolve and get refined. And you know we have seen that in other services and you know the user contribution that’s going to be our contribution to the world of search. And you know it will take time to refine but we are really happy with the results we are seeing so far. And you know the ability for users to have some say in what the search results are I think it?s a powerful thing. But I think that totally fits with every whether you know so many of the trends that are happening on the web right now is the ability for users to impact and have some input into what they see. And also at the same time to kind of help others in the process of you know of changing the results and you know voting for the things that we really like and getting rid of things they don’t I think it?s very empowering.

Well there are two ways. First of all the amount and I can partially go down the technology works and you know not completely for obvious reasons. But the people rank technology has built into it. An analysis of how people behave in such a way that if they seem to have intentions of like elevating to the top websites that you know that perhaps aren?t legitimate for the search that?s been conducted then it?s going to spark that and it?s not going to let them have as much influence on it. Secondly if a user sees a site that they really don?t think belongs in the result we give them a block button which they can use to remove the result. That can be effective immediately removing it for them. It does casting a negative vote for that site which we will include in our analysis.

Yeah it wouldn?t be, it would not be shown within future search results.

Well we use an advertising base model. Currently, we are using Google AdSense which is a great product for search.

Well you know advertising is sort of like the underlying fuel of the internet today and I think that works because there is actually a large amount of commerce really happening unlike the bubble internet base when there is a lot people advertising hoping that that would you know get them an idea instead of something. Today I think a lot of the underlined business is going out, the advertising, they are seeing a return on their advertising dollars and in fact I think there is more demand for good ads than there is inventory available and so I think it?s a very healthy market. And I think especially it works in the case where people are searching because it?s contextual and that you know so we see it as a stable and reliable revenue source for the future.

Okay well yeah good you know it?s quite of touching on that, so let me expand a little bit. So looking at the, what it looked like you know say in 1998, 1999, 2000 there is a lot of advertising money that was in the market and that was coming from n number of different sources. It was coming from large companies who already had existing ad budgets experimenting with some of that ad budget online and then it would arrow for the big automakers, to the telecoms, and you know software, and computer companies. And then you had Internet companies who had fractionally invented venture capital and they were hoping to get an audience faster than their similar partner, you know similar companies you know so you had packed food online, you know 3 or 4 of them each of them trying to outstand the other to get market share faster because it is just believed that whoever got the market share first would then dominate the market. And then you had also advertisers who had a product to sell, you know the Amazons and others, who really were selling online and we were using the Internet as a ways to find customers. You know of the 3 what happened I believe was the highly speculative venture-backed ones ran out of money and you know had various events, being sold off, shut down whatever and obviously their flow of cash stopped and some of that was very significant in the end, people spending 10, 20 million dollars in ad campaigns. It had the major world?s leading advertisers and consumer products basically back out of the market for a couple of years. You know it was no longer attractive to be part of the whole Internet and I think the 3rd category of the Amazons and eBays and so on they maybe scaled back it but they stuck in there. You know they continued to operate as businesses and what I might think was you know after one had had some time for some healing it was in 2003-2004 people started to come back into the marketplace. I think Overture showed that if you could generate, if you demonstrate a return on investment that if you only had to pay, if you saw some of them yield then other people just came and visited your site then people would at least try that because you know there is a risk in that, a very little risk you only pay if people visit your site and adds to your objective. Some number of those people would become customers and so advertisers began this Overture and about the same time Google was looking around for a business model, till that time they hadn?t actually served any ads. And they picked up the model and ran with it and did a great job of building AdSense and I mean AdWords at the time and so we now have you know Overture and AdWords as kind of two viable advertising vehicles. I think over the past 2 or 3 years more and more advertisers have realized there is a great way to generate business to the extent that is now being very lifeblood of many web businesses whose vast majority of their customer attraction came down through one of those services. And then on the other side of the coin you got web businesses reformed who are now providing the inventory that the pages on which they advertise as almost their sole business model and so you have got a marketplace.

There will always be more than one service.

That?s true like I think you are going to see competition and well let me put it this way, I think like AdWord has done a great job and they are successful because of that. I think if anyone else also does a great job, they will be successful to that extent, now I mean it?s clearly there are market effects and you know certainly you know they may attract a larger percentage of it. I do not believe it?s a natural monopoly or even monopoly is out the right word but eBay has a very strong market because anyone buying or selling something needs to kind of be aware of all the buyers and sellers are. And there is a little bit of that effect here. But as an advertiser I don?t really care if all like the buyers are in a particular marketplace where I buy my ads. I just want to make sure I get proper coverage. And so to the extent that I can reach my audience I will buy on any network I mean an advertiser is not below to NBC or CBS. They just know that NBC has a great show that they want to sponsor maybe they would like sponsor Fox for 24 because they think that?s got the right audience for them. So I think an advertiser will advertise wherever they can find their audience. And to the extent that there are a variety of audiences out there I think will see multiple networks. And I think there will more of I mean that said definitely nothing wrong with Google, we use it very happily and probably we will continue to do so unless something very compelling came along then I guess we would change.

Yes, I do believe that?s true. I think sort of a, almost I am often in conversations with people now whether they almost chuckle because when they see a business model evolving that it?s so similar to when they saw you know five years ago. And I think there were a lot of really great business models that had already evolved, people thought of doing these things. It?s for almost every new internet company now I could think of a counterpart that existed five or six years ago. You have del.icio.us and other social bookmarks now they are with Hotlinks and Backflip, you know six or seven years ago for social networking. You know you always have Friendster and MySpace there is now, you had you now SixDegrees and others back then. So the business ideas and business models were there the service ideas but you are right the philosophy was very different. And I think what happened was for a variety or reasons people companies basically over invested in growing their businesses and I think in some cases obscenely so. And its kind of the trap they are in and if they didn?t expand quickly and grow they would quickly be dwarfed by someone else who was willing to spend that way sometime with no stable future in mind, just spend someday enough revenue would come and if you look at the multiples people were spending as so much money in achieving valuation so high that you could never reasonably expect to sell that much dog food or whatever other product to justify that valuation. So it was basically a house of cards that couldn?t possibly sustain itself and the problem was the public markets were even complicit. It would be one thing for you know venture capitalist to over invest their money and then you know reap the rewards for their decisions. But you have the capital that public markets were complicit so these companies were actually able to get access with these financials and you know weak financial underpinnings and which you know made the whole thing extremely unhealthy.

Well, let me say that the philosophy has changed significantly not just for us but for the whole market is that I think we have got a lot of the same business models and same business ideas and number one you have got a robust advertising market as I said earlier that has this underlying commerce and real revenue happening. So that?s the first thing. The second thing is all of us have a much more circumspect approach to things that says look we are not going to spend way out ahead. We are able to do business on small fraction of the investment. The cost of the equipment and the capital is much lower. We are not paying ridiculous amounts for rent and other services, and so we are able to spend money carefully and build legitimate businesses with the idea of quite reasonably having a cash flow in a relatively near future. So what I see for Wink is that we will build our business, we will attract users who will find our service satisfying. We will earn revenue form the advertising and we will be help to build a sustainable business that earns revenue for the service we are providing and provides a real grade service for people.

Well, I think what we have done today is actually have added Wink to other models at this point, so you know, we are using you know Google provides world class search results and having used several of those models and what we have done is added Wink to that model. The additional models I think we will pick and choose which ones we believe can make a difference for our users and that is always open, you know we are always open to adding new techniques as we understand them and I think a lot of them alls have been tried and haven?t delivered and so those are on hold, you know, they have been tried by different companies over the years and you know, we probably have to be convinced before returning back to some of those, but I am sure there is new things that you know we haven?t even done yet that would possibly be very interesting.

Well I will tell you some of those insights are constant throughout and some have very much been different for the time period, you know, what was right, or what you needed to do in each period during which I have been active as an entrepreneur has been in some cases wildly different. And so some of the things that I found are consistent throughout, or that there is a real leadership and team are really fundamental, I mean, you can have the greatest technology in this business of idea but I think you need to build a team that is, a team that works well together and handle our custom and respect for one another and we have been fortunate in each of the companies that I have been involved in to have you know our teams that really could count on one another and you know they are talented, energetic, you know, working integrity. So I think, I will always put that at the beginning of the crux, because you can solve any problem with the right team. And then you are around that you have got to build other things like, I have always leaned towards having a strong technical foundation. You know it is probably a little bit about sort of when my roots come from having been an engineer and then moving over to the business sites so I tend to get into businesses where there is a strong technical need you know and search certainly is one of those in the past without forcing with you know is advertising or ad serving. And then what I have also found is that there is sort of an intuitive web of dependencies that if you try to build a company quickly you will find you really have to sit and understand these interdependencies. For example, you know you need credibility in order to raise capital. We also need credibility that in order to hire people since you are convincing new people that this new start up can be successful and that?s the team you need to actually build the product and in a way it?s building the product that gives you the features and the benefits that your will like. The users have to decide that traffic that increases your credibility so you are trying to have to get all of these things going in sort of a virtual cycle if you are trying to build a company quickly. So I think that?s been very consistent. I think some of the things that have been kind of different from time to time instead of how quickly you spend your money you know when to use it because the opportunity is in front of you and when to conserve it in case you are going to need it in the future you know certainly today a lot of companies are using or making great use of you know low cost solutions like using a lot of open source and using you know really efficient computing platforms and so on. So actually in some ways a great time to start a company because you have got all of these resources available to you and can really get something going quite quickly.

Oh no question I completely agree with Joe on that one. When I was AdForce we invested over 20-million dollars in hardware to serve ads. Now we did serve an enormous volume of that. We were serving 700-million ads a day when I left and you know and the business was definitely you know it was paying off I mean we were basically about a cash-full at the time we got a credit by CMGI and but if you look at you know so that was a business model where you know a kind of you know just barely worked in terms of you know being able to serve enough ads and to make enough revenue to justify the capital and the cost of the team. Three things have been reduced, three things have changed and you get to argue that I would say that the cost of the hardware itself has come down by a factor of 10. The capacity of the equipment has increased by a factor of 10. And the cost of development has dropped I don?t know by a factor of you know 2 or 4 which is based on using a lot of reusable code and open source and things like that. And so that?s you know that?s like a 200 or 400 fold difference over you know your cost per transaction of whatever you business was back then.