3 ms·

The value of Google Maps directions logs

54 points by joshfraser 16y ago

boscomutunga 16y ago
Yes,i think that kind of data is very valuable. Another application of this kind of data is in transport and movement statistics of a population.
ChuckMcM 16y ago
This is a good example of a meta-data injection. Not clear yet what the outcome will be on the overall system. Kind of like auto-complete for places. The scary thing will be creating need feedback loops. So club A tries various marketing strategies and compares that with its rise (or fall) in the 'destination intent' index. These then trigger more specific marketing but then carves out a population which is responding to the messaging. Combined with groupon style offers its going to be an interesting couple of years.
thaumaturgy 16y ago
Well, I guess I'm fucked for the future. I'm extremely skeptical of the value of the "wisdom of crowds". On social sites, I find myself often finding value in overlooked comments or articles and, worse yet, often finding little value in the most popular comments and articles. On technical sites, popularity again has little to do with helpfulness, although this has become probably the single greatest ranking metric in search results. For one self-promoting example, a year and a half ago I posted a fix for an annoying osCommerce 3.0 bug (http://robsheldon.com/oscommerce3fix http://robsheldon.com/oscommerce3fix); it never showed up in search results for queries like "oscommerce HTTP_X_FORWARDED_FOR". That doesn't really bother me, except that every single time I'm banging my head against some obscure problem, in the back of my mind I'm wondering how many little blogs there are out there that have posted the exact solution I'm looking for, except I just couldn't find them if my life depended on it because they aren't ... popular. As an avid hiker and former climber, the last thing I want to do if I'm in unfamiliar territory is be directed to wherever everyone else goes. Hell, I don't even usually want that if I'm doing the tourist thing in some city. Ugh. I'm starting to develop this really powerful sick-to-my-stomach loathing for "crowdsourcing".
- defrost 16y ago
  So you're saying you've a keen interest in the inverse of this data set then?
  - thaumaturgy 16y ago
    No. Saying that I don't see value in a particular thing does not mean I see value in its opposite. If we must measure and rank the value of everything, I'd prefer it to be done on some metric unrelated to popularity.
chrismealy 16y ago
The places I most want to go to I don't need directions for.
- ZoFreX 16y ago
  That seems to be the biggest flaw for me. There are shops that I go to every week, and I never type them into Google (Maps or otherwise), and I'd certainly never need directions. If I need directions to "B", the only thing it's a good indicator of is that I probably haven't been to "B" from "A" before. Value could only be determined if I went there more than once, but in that case, I wouldn't "tell" a search engine.
- ratsbane 16y ago
  Having all aggregate phone location data available for comparison would be useful in balancing that. If you could compare where people are to where they're searching for you could figure out what new places people go to and what places they go to but don't search for. Rank places based on how often people go back there - sort of a geo-Backrub. Google and Apple obviously can have that data, but do privacy constraints prevent them from using it?

54 points by joshfraser 16y ago

Yes,i think that kind of data is very valuable. Another application of this kind of data is in transport and movement statistics of a population.

ChuckMcM 16y ago

This is a good example of a meta-data injection. Not clear yet what the outcome will be on the overall system. Kind of like auto-complete for places. The scary thing will be creating need feedback loops. So club A tries various marketing strategies and compares that with its rise (or fall) in the 'destination intent' index. These then trigger more specific marketing but then carves out a population which is responding to the messaging. Combined with groupon style offers its going to be an interesting couple of years.

thaumaturgy 16y ago

Well, I guess I'm fucked for the future. I'm extremely skeptical of the value of the "wisdom of crowds". On social sites, I find myself often finding value in overlooked comments or articles and, worse yet, often finding little value in the most popular comments and articles. On technical sites, popularity again has little to do with helpfulness, although this has become probably the single greatest ranking metric in search results. For one self-promoting example, a year and a half ago I posted a fix for an annoying osCommerce 3.0 bug (http://robsheldon.com/oscommerce3fix http://robsheldon.com/oscommerce3fix); it never showed up in search results for queries like "oscommerce HTTP_X_FORWARDED_FOR". That doesn't really bother me, except that every single time I'm banging my head against some obscure problem, in the back of my mind I'm wondering how many little blogs there are out there that have posted the exact solution I'm looking for, except I just couldn't find them if my life depended on it because they aren't ... popular. As an avid hiker and former climber, the last thing I want to do if I'm in unfamiliar territory is be directed to wherever everyone else goes. Hell, I don't even usually want that if I'm doing the tourist thing in some city. Ugh. I'm starting to develop this really powerful sick-to-my-stomach loathing for "crowdsourcing".

defrost 16y ago
So you're saying you've a keen interest in the inverse of this data set then?
- thaumaturgy 16y ago
  No. Saying that I don't see value in a particular thing does not mean I see value in its opposite. If we must measure and rank the value of everything, I'd prefer it to be done on some metric unrelated to popularity.

chrismealy 16y ago

The places I most want to go to I don't need directions for.

ZoFreX 16y ago
That seems to be the biggest flaw for me. There are shops that I go to every week, and I never type them into Google (Maps or otherwise), and I'd certainly never need directions. If I need directions to "B", the only thing it's a good indicator of is that I probably haven't been to "B" from "A" before. Value could only be determined if I went there more than once, but in that case, I wouldn't "tell" a search engine.
ratsbane 16y ago
Having all aggregate phone location data available for comparison would be useful in balancing that. If you could compare where people are to where they're searching for you could figure out what new places people go to and what places they go to but don't search for. Rank places based on how often people go back there - sort of a geo-Backrub. Google and Apple obviously can have that data, but do privacy constraints prevent them from using it?