Mobile Google...idea #38/100...
Perhaps you have tried mobile google. You can try it via your regular browser. It's instructive to search for your own site and then see how the Google transcoder copes with rendering to a mobile device.
Search results from mobile Google are ranked the same as standard searches. This is problematic because it would be better to take into account the "viewability" of sites via the transcoding process. Some sites are clearly not that viewable and best to eliminate them from the results, especially given the high cost in visiting sites on the mobile, both in time and most likely money (data tariffing).
Of course, the search engine and transcoder don't know which sites are viewable without attempting to visit them, because transcoding appears to be done on-the-fly. I guess that there's no reason, other than cost, why the search indexing process couldn't include transcoding, at least to assess the viewability.
How would a transcoder assess viewability? Well, there are probably a number of design rules that could be formulated to detect the likely success in rendering to a small screen, although it is not a straightforward problem.
Therefore, it seems a sensible idea for Google to publish some design guidelines to let site designers understand the transcoding process. This would hopefully allow, and possibly encourage, some site designers to start thinking about the mobile experience, something that hitherto has escaped the attention of most designers and related tool providers.
It does seem that for the foreseeable future, mobile access to many sites will involve some sort of transcoding process. There are already a variety in use, not just at Google. However, the use of Google's transcoder is obviously of paramount interest, as so many users access the Web via their front door.
Google, and any other transcoding front-door, could go a lot further if they wanted to. It seems in their interest to do so, not least because if they can offer a viable mobile access service, then they will attract more traffic into their ecosystem and money-making machine. Mobile adverts, especially with the latest ad-links approach, are a possibility. Let's defer the advertising conversation to another time...
We have a wonderful convergence on a common language approach for mobile websites, namely XHTML, or similar (i.e. the -MP variant). However, we have very little convergence on a common "design language", or "grammar", that combined with transcoding approaches will produce a consistent and usable mobile Web.
Transcoding can be vastly improved by some kind of site annotation to guide the process. This is an activity in the W3C, but with little tangible output at present (not that I've seen).
A long time ago (2002) I posted an article suggesting that operators should host an annotation system that any site designer is free to use (at least via the operator's pipe). I suggested the idea to someone at Volantis, who make a site-adaptation product (which strictly-speaking is not transcoding).
I suggest that the Google team should make a lot more effort in this regard. If not them, then any search-engine company that wants to attract more mobile traffic to its site. Not only will it result in a greater user experience, but it will improve mobile search efficiency. Effective mobile search is of vital importance to the mobile Web.
Join my email list
Subscribe to my "100 Mobile Product Ideas" free e-book