Eurekster is looking to provide custome search engines to small and large content sites. Enabling both Eurekster and its partners to monetize the search results.
Interview conducted by Nathan C. Kaiser on Thursday, November 17, 2005 in San Francisco, CA.
We see search as being one of the last things to be opened up on the Internet. The Internet is becoming about personal publishing. First people had the power to publish web pages, blogs and podcasts. We enable people to publish their own search engine. The reason it is a publishing platform is because you have all this content available on the Internet now people can control what is presented and how it is accessed. Each of our search engines is completely customizable.
Currently there are three large players within the search field. You go to one of these top providers and type in something. Everyone else that types in the same term or keywords will receive exactly the same results. We allow our partners to create their own bias of the world and in effect customize the search results for their area of interest.
If a woman is pregnant and visiting a medical website or a community site focused on pregnancies, when she types in terms such as vitamins, health, etc. she is looking for very specific results. Typing those keywords into any of the three main search engines will provide all types of results.
The large search engines to some degree provide results that are tailored to the mass market, not to specific needs or interests.
We think that search will become like all other forms of media, where you start with one large player who tries to be everything to everyone; like radio at its beginning, the large TV networks, etc. What happens over time is that you get a splintering of lots of different providers specifically tailored to small and niche markets. We have seen this with the explosion of cable channels such as Nickelodeon and the History Channel.
Rather than having one view of the world through one search engine that determines what everyone should and shouldn’t see we believe that no one is smart enough to do it accurately. We empower our partners to create a view of the world that represents what it is that they think is important. Normal search engines do everything possible to ensure that the end user can’t affect results. We totally reverse that by giving the end user the ability to promote, demote, block, include, rate, and review.
A concern that is voiced about this approach is that people could create SPAM search engines. Our retort to that is that they can do the same thing with web pages, and people simply don’t go to those web pages, and they won’t use SPAM search engines either. If the search engines are good, then people will use them. People who are passionate about architecture, who build a web site about architecture, or blog and podcast can now build a search engine around architecture.
A publisher such as nPost.com could create a search engine that is specific to entrepreneurship and startups and we would be able to control that search. So we would have control, but what about the end user? How do they interact with the software?
Our goal was to try and capture the wisdom of the crowd that visits a certain website. People visit a website or blog because they have a certain view of the world. We try and capture that actively and passively. The main way is passively, simple because people tend to be lazy and don’t want to work. Every time someone types in something to our engine and clicks on a result we then learn from their behavior and modify future results based upon that data set.
The reason we chose Swiki as the name for the product is that it is a combination search engine and wiki. Basically, it is in the power of the community to go in and change things as they see fit. At the moment we have it set up so that only the moderator can actively add, remove and prioritize different search results, but in later version it will be open to all users. Publishers will also be able to have different levels of control.
Do the individual learnings that happen with the smaller publishers roll up into the larger service?
They are kept localized, because we believe that what is important for one group is not important for another. The search results can be quite divergent from one another based upon where someone is searching. It is the idea of having lots of different views of the world. We know that vertical searches are happening more and more. If you want to search for travel options or jobs you end up at a vertical search engine. Search has evolved from computers deciding what was relevant (Alta Vista) to Yahoo’s major innovation which was having people identify the top sites within a directory and finally to Google’s link analysis. The next step is to expand past what we think is relevant to what the end users think is relevant.
The actual economics of it are quite dramatic. You have a multi-billion dollar pay-per-click search market that is dominated by a handful of players and there are tens of thousands of web communities that are made up of valuable audiences. At the moment they are missing out on any revenue from the pay-per-click search market. They could put up a hosted search by one of the larger search engines, but that would only provide the same results as what everyone else is receiving.
What drives people to the community sites, and will entice them to go back once they click away via your search engine?
If people want to do any searches based around entrepreneurship, they will go back to nPost.com. They go there to read the articles, learn more about entrepreneurship and to search the specific engine that has been developed on nPost.com. The real value is that search traffic to your site generates revenue, and our model shares all revenue with our partners. Like all other forms of media the more targeted your audience the more valuable the advertising. A cost per view of a newspaper ad is a lot lower than an ad within Golf Weekly or Bankers Digest.
Ultimately, in the long term people will not only continue to be able to buy keywords but also keywords targeted at different audiences. nPost.com has built up a loyal user base and that base has a demographic that is more valuable than a generic audience.
We have two levels. One is the swiki product, which is very much a self serve model. Then we have our larger SearchPublisher software, where we go out and provide in-depth customization. We are going after websites that have a lot of targeted traffic. A lot of them already have relationships with the advertisers that are important to their end users so they are really keen to see if they can make money from the search engine part of it as well.
That is something that we are still learning. We tend to start people with the swiki so that they can quickly launch with a search solution. Then based upon the traffic and performance we can offer greater customization and a complete look and feel that matches their site. We feel that this approach works for the smallest blogs as well as large content sites.
Absolutely. Google has proved with AdSense that making at least some money is quite motivating. A lot of people have been developing their sites and communities for quite some time and now they can monetize it. Previously they had seen this large profitable area of the web pass them by, but now they can tap into their proprietary community and leverage the pay-per-click model as well.
The company started 7.5 years ago when we had a company called Global Brian that we ended up selling to NCBi during the dotcom boom. Then NCBi was trying to compete against AOL, Yahoo and others and ended up closing down. We were able to buy the technology back and began targeting it with a company called SLI systems. We were seeing what was happening in the search market and what was happening in the social networking market. With the whole approach to the web 2.0 where the goal is to give the power back to the people that it seemed like a logical next step.
It seems like a major transition from top down, with large publishers providing content to the end users to a much more bottom up approach, with users creating some content as is the case with WikiPedia to determining their own search results. This has the potential to completely change the current model, in some ways for the better because users can become much more involved.
Some publishers get it way faster than others. Publishers want to get people to their website to view their content and they don’t want people utilizing a search tool because they inevitably leave the website. Other publishers admit that it is a good business model as it has worked very well for Google. Their whole aim is to push people away from their website as quickly as possible. If you have a website you can either manually pull all the content to your site or you can point to all the other relevant content out there and make money. It is much more scalable.
As one of the co-founders how did you transition into the role of Chief Scientist and exactly what is that role?
Good question. I started off as the CEO, but it quickly became apparent that as I am based in New Zealand we needed someone in Silicon Valley who could put together the team and do deals. All of our sales and marketing is in the US and all technology and product development down in New Zealand. The reason for the title is that I am not technical enough to be the CTO and don’t actually write any code, but I do have ideas and a vision. The title was bestowed upon me because we couldn’t think of anything more relevant.
There is not one way to be a successful entrepreneur. It is very dependent upon your technology, your skills and your ability to recruit others. You do have to be able to attract people who can do the things that you as an individual can’t do. No one person can do it all. With Eurekster, we believe that our future lies in partnering with another provider that has the other parts of the solution.
One of the key takeaways is that you have to be very open and flexible to how you approach any new business.
You can’t actually know what the most important thing is until you get it out there. It is vital to get something in front of end users as quickly as possible to get their feedback and ensure that you are going in the correct direction. A lot of good ideas emerge from putting them in front of end users and seeing what they think, and not necessarily from your own brainstorming.
A previous company Real Contacts was focused on how people actually hire. We took a long hard look at how we had identified the people that made the most contribution to the team and it was completely by chance. Someone happened to mention to someone that happened to be looking that we were hiring. The reason that these networks are important is that people who tend to recommend each other tend to have a lot in common, and a similar view of the world. It is their approach to the world, their attitude and persona that is more important than individual skills. You want people who when presented with a challenge say “how hard can it be?”
About nathan kaiser
