Friday, 24 January 2014

Search, Optimization and the Personal Map Experience

Some idle chat over lunch at the DGI conference yesterday got me thinking about personal maps. By personal maps I mean the digital maps we use every day on our laptops and smartphones; Google, Apple, Open Street Map and the rest.

Obviously the huge draw of these maps is that they are interactive and searchable, not only can you find where you’re going but you can find your nearest coffee shop, railway station on pub. Or rather you can find the nearest one of those that the map provider has decided to list (for whatever reason). I used to work for what was effectively an SEO agency and as a result I find this ‘searchability’ and the personalisation of results very interesting. Far from being (as some critics originally suggested) a uniform and boring representation of the world these mapping applications deliver an incredibly varied user experience. Obviously there is a debate to be had about the amount of influence commercial bodies have over what the majority of people will regard as a true representation of the world (which is the reason for Open Street Map).

However I want to take a simple ‘out of interest view’ on the personalisation of user experience that exists with the modern digital map. We know that Google (and the other search engines) personalise results based on previous search history and a range of social metrics as well as providing auto complete suggestions that are an amalgamation of popular searches and search history. This seems to be similar for maps with two major differences. The first is obviously the location factor, if you are searching Google from a desktop it can make a rough guess at your location based on IP and will localise searches based on geographic terms you enter (e.g. pubs in Putney). If you are using Google maps with a GPS enabled device however this becomes much more specific to the extent that a more nebulous search (local pubs) is likely to return relevant and accurate results.

These results are then of course affected by the second difference; advertising. On a normal Google search page pay per click advertising entries are (relatively) distinct from the organic search results. In Google maps this is somewhat less clear and the ordering of results by relevance (A to whatever) does appear to be impacted by this. For example when I search ‘local coffee shops’ from my laptop in West London result A is in East London about 7 miles away, not really local for a city! (a lot of the other results are quite close though)

Anyway the sum of all this is that two people standing in the same place searching for the same thing on Google maps (assuming they are not browsing incognito) are likely to get very different results or at least results in a very different order. Now this got me thinking of an interesting experiment/gimmick/probably impossible business idea...

If what you see on your map is determined by where you have been and what you have searched before then what if you could swap with someone else’s history? Ignoring the obvious data protection issues for a second this could be quite interesting. Say I go to Manchester for the weekend with some friends, I don’t really know Manchester that well and like most people I am quite lazy when it comes to researching things. As a result which restaurants I eat at and which bars I go to are likely to be heavily influenced by what comes up to on a Google maps search (pending no terrible reviews). Well what if I could download someone else’s personalised search (not the history just the algorithm results)? Then I could end up visiting a completely different set of places that I would never have been to otherwise.


It’s like a horribly impersonal 21st century version of letting your friend’s friend who you’ve never met organise the trip, it might be fun though. With the added bonus that you can’t yell at a search engine for suggesting a rubbish gig. 

No comments:

Post a Comment