I had a great discussion with Tomi Poutanen last week about the future of algorithmic and social search. Tomi identified 3 challenges faced by algorithmic search:
- The size of the web they are searching and indexing
- Subjective queries, such as “what’s the best hotel in New York” can’t be addressed by algorithmic search
- The spammers use algotihms too, and there is an inherent arms race between spammers and search engines
But Tomi does feel that algorithmic search is still better for navigation and deep research. So, ultimately, there is a home for both.
We also talked about the return of tagging, and I asked Tomi why tagging would work this time, when the metatags of the past were such a dismal failure. It turns out that there is one key difference. Metatags were used by webmasters to indicated what there site was about, and hence were highly subject to SPAM.
But with social search, tagging is done by people to help them bookmark content they like. On sites such as del.icio.us and Flickr, the most heavily tagged content rises to the top. So it’s a bit like an election process. It becomes very hard to spam this, because spamming of other people’s tagging decisions is difficult to do algorithmically. In addition, if a spammer succeeded in influencing how their site ranked for one type of tag, the return just is not that high.
For more, check out the full transcript of my interview with Tomi Poutanen.