The future of search

I ran a great Churchill Club event the other night (I say modestly) on the future of search engines, asking questions around “are search engines still effective” and “where is the money being made tomorrow”?

Our panel was:
Angela Beesley Starling – chair of the Wikipedia foundation, co-founder and vice-president of Wikia;
Alan Long – research director, Asia Pacific for Hitwise;
Tim Giles – director, Sagital Media.

The results of the conversation were fascinating, so I thought I would pass some points on.

The mega trends around search engines

Search is becoming more complex as users are entering in longer and longer search terms to find more specific information (think “cheap flights Melbourne Brisbane”, rather than just “cheap flights”).

This tends to indicate users are embracing the long tail – less prepared to accept generic results. Which creates a problem for search engines as their current algorythms tend to promote sites with lots of links, which is not necessarily what users are looking for.

Human intervention is desired to make sense of the overwhelming masses of information, however crowd sourced search doesn’t seemed to have worked though. Wikia had to close down its Wiki Search, and Google’s Search Wiki hasn’t had mass appeal.

But websites with human filtering are on the increase. Filtered video websites such as Wonder How To? for instructional videos or Political IQ for politics are becoming more popular than simply searching YouTube. Wiki Answers after three months was getting 10 times more visitors than Wiki Search after two years.

So what do companies need to think about? Of visits to Australian websites, 32.4% are from search engines (of which 95% are from organic search results, not paid links in search results). This means that search engine optimisation should be a much higher priority than just buying the right AdWords.

Social networks (Facebook, MySpace and the like) generate another 11.6 % of traffic to Australian websites so they shouldn’t be forgotten as the next highest source of traffic.

It should also be remembered that search engines are driven by the best way of monetising their search results, which is slightly different from actually giving the best results. Therefore it’s possible that anything that improves profitability by skewing results could be implemented (remember it’s all black box stuff).

Because of this “increasing long tail” of search, organisations should be looking much more closely at their cost of acquisition and be careful of their AdWord and SOE spends. Budget could well be put to better use optimising content for being picked up in a small number of organic searches.

Companies should also remember that just because the world wide search for “widgets” may be one million users, they may only obtain sales from Australian searches, which could be a fraction of the former- that is, the internet may be global, but your customers are still local. Don’t get suckered into big spends.

Well, what are users up to?

Search engines are still the number one tool on the internet. Last year Australians went to search engines 12.5 times for every 100 pages visited. In NZ this rate is 12.2%, and in the US 9.4%.

Traditional search engine use is becoming a more complex landscape.

Search terms are becoming longer, but the chance of having a successful search (clicking on a link rather than re-searching) is reducing. Last week Australians entered 3.6 million phrases into search engines.

The length of the search term is increasing as users become more sophisticated. Last year, use of search terms of:

• 5 words increased by 7.5%.
• 6 words increased by 10.5%.
• 7 words increased by 14%.

The top 10 words searched for by Australians last year were all brands (Facebook was number one in Australia) and Whitepages was the only traditional brand in the top 10. In fact you had to go down 23 places before you got a non-brand search term.

Where are the emerging opportunities?

Interestingly, two new (complimentary not replacement) search areas were discussed, both with different ways of monetising the results.

• Realtime search – Many organisations are now monitoring real time conversations on platforms such as Twitter and dropping in to offer recommendations when their search terms are matched. They are also monitoring conversations to understand trends and context when their search terms are used.
• Non-text search – Shazam is an iPhone application that searches for details on songs it hears. When the song is identified, the user is offered the option to purchase the song at the iStore, and it actually works really well. Apple is leading the way when it comes to monetising non-text based search.