There is a good post on Techcrunch today about the impact of Google Instant on SEO. SEO company RankAbove has done some research and from the two weeks of data available it seems that one of the big changes is reduced traffic to long-tail keywords, here’s why:
Because Google Instant focuses the user’s attention on more popular search phrases, by suggesting them and showing their results as the user types, fewer users will use longer tail, less popular phrases. Searchers typing a longer, more obscure variation of a keyword are unlikely to complete it before they see the results—automatically loaded onto the SERP [search engine results page] —they were looking for.
A company selling iPod Car Adapters, but ranking for a lesser-used phrase, like “iPod Car Cables,” is unlikely to get much traffic from that phrase, as a searcher using Instant would almost never get to that search before seeing relevant results. For a site counting on unpopular, “forgotten” long-tail phrases like that, their traffic will dwindle.
I think startups are likely to be disproportionately hit by this development, losing out at the margin to companies that have the scale and resources to optimise for the most popular queries.
The other change of note is that less people are clicking through to the second page of results, which will similarly lend advantage to companies that have the muscle to get onto the first page.
Caveat: This analysis is based on just two weeks of data and the results might change as volumes increase and people get through their periods of learning to use Google Instant and settle down to a consistent pattern of behaviour. The other thing is that Google Instant doesn’t work if you search directly from the nav bar or toolbar, so any changes wrought will only apply to a percentage of users.
Update: In the comments Benoit Cundy pointed me to this chart which shows that very little traffic went to the second page of results anyway.