> Use of the Google Static Maps API is subject to a query limit of 1000 unique (different) image requests per viewer per day. Since this restriction is a quota per viewer, most developers should not need to worry about exceeding their quota.
>>
At $work in Australia, we went through this a couple months ago. After an initial quote that was beyond ridiculous, we worked out a much better deal with the local Google reseller. In addition, to help keep map views down we stopped unconditionally loading the map. Instead we have a placeholder image which when clicked loads the maps API dynamically. Google's fine with this behaviour as long as we don't use their map image.
We looked at other map providers, and in particular Bing and TileStream/OSM. But Bing's cost was 1) nearly what Google wanted and 2) confusing to try and get an actual price because there are about 10 different ways to get a Bing maps license. OSM ended up being ruled out because 1) we weren't confident in the data quality in India and Malaysia and 2) I couldn't find an easy way to geocode.
I wish I had known about the Mapstraction project freyfogle mentioned. I would have seriously considered using it since we were already moving from v2 to v3 of the Google Maps API. Future proofing our maps implementation would have been nice.
(full disclosure; have patches in many mapping frameworks, and a surplus of opinions)
You're probably better off just using a straight-up real maps API, like the MapQuest open tile API [1], and an open-source framework directly. And for this, use Leaflet or Modest Maps javascript. Not OpenLayers unless you really, really need OpenLayers.
There is also http://openlayers.org supporting all different kind of formts (GeoJSON, GPX, WKT...) and different providers like google, bing, osm, ... It even supports mobile
As someone who runs a service where a google map is THE interface, I don't think these rates are all that bad considering how powerful of a tool gmaps is. However, my service doesn't come close to the 25k required before billing comes in to the picture.
I do wonder what the people behind padmapper.com think, which most likely will get charged and has a similar gmap as the service interface.
The really interesting cases will be things like Yelp, where the map is a small feature to the overall service. At 1.5M page views/day that's $1,500 per day...
Congrats on PadMapper - it's a fantastic service, and makes looking for an apartment in a new city infinitely easier.
This doesn't help much in this situation, but thanks for all your work on it; I can't think of a single person in my social circle who's moved recently and hasn't used it.
You're welcome! And, I'll keep that in mind :-) I'm almost certainly going to leave the favorites and everything else there free, though. I'll come up with something.
It would be fairly ironic if Yelp has to pay, seeing as how Google uses Yelp ratings in their maps freely and wants to charge Yelp to use maps on their site.
Yes, it is just too high. I use a v3 styled map that logs about 15K map loads per day (there is a map on just about every page). That means my total cost per day would be $52, just for maps. I would have to make about $3.5 CPM via adsense to break even on map payments (that doesn't even include my hosting costs and other business expenses).
This change is going to force me to revert back to un-styled maps and just hope my traffic doesn't grow above 25K. If that happens, I'd definitely switch to another map provider.
I think the key thing here is that the google map is a backdrop for my website/app, it is not the main feature. I am happy to pay a reasonable price for my usage, but I am not going to be forced to pay an unreasonably high price for it.
It's been mentioned elsewhere in this thread but if you are using the client side maps js without an api key how do they track usage? Maybe there is something we are missing?
Yes it is, at least for us. We run a 15 mio pageview/month website, completely reliant on advertising, and we have a google map on probably 4 mio of those pageviews.
So our bill is going from free to 16.000$ pr month
$4 cpm is interestingly enough the same as we charge our own advertisers leaving zero profit margin for us. We are now in the process of converting our website to another map provider.
I wonder if any Google reps are reading this? I highly believe that this forum represents the general feelings held by most developers using Google maps.
Google appears to be forgetting that their success has more to do with their name at this point than their products. Not to knock Google Plus, but, had it not had the name Google in front of it, it would have been just an afterthought that would have received a half-pager from Alexia on TechCrunch. Making Google Maps free is what led to Google Maps being ubiquitous and that eventually led to Google Maps becoming the defacto map provider in the same way that Google Search is the defacto search provider. Now, even my mom tells me to "Use Goooogle Map". This type of tactic keeps Google on the tip of everyone's tongue ... even the tongue of my computer-illiterate mother.
Although I avoid being an alarmist, this recent change will lead to developers shying away from Google Maps in their own sites and/or switching away from it if they've already used it. This will in-turn expose end-users to other experiences and will gradually reduce people's association of Google to all that is "finding stuff using the internet".
So, yes, what I am implying is that Google should run Google Maps at a loss for the simple sake of maintaining their brand's presence in the online world. Some may say that this is bad business since it involves losing money, BUT, I say it's quite the opposite. Having my own mother be aware of Google Maps without ever having used it herself is testament to the success of the brand.
I am confused about this too. It seems like 99% of sites using maps would be doing it client side in which case the cost change wouldn't be a big deal, but maybe I am missing something.
They are talking about client-side APIs, but the limits are not per client IP address. They can look at the referer header to see what site embedded the map.
Google blocks you from seeing referred search terms and you flip it around so they can't see what server the requests on their API are coming from. Seems kind of fair, right?
So you think that the Geocoding API will work via IP limits as described in the link I provided, but regular Maps API calls will be subject to domain limits?
Is this for public facing sites? I remember there used to be an annual fee of $8 000 plus a pageview hit if you had an intranet app. We ended up using OSM because of it.
Google charging for this is good only in that it guarantees attempts by other parties to create independent solutions that are lower cost.
Sometimes I worry google is too good at things like gmail, maps, etc. and it causes others to not even try to develop competitive services
I was surprised to discover bing maps is far better than google maps at some things, particularly their "bird's eye view" and they find addresses far better in my town where Google will pick the wrong spot.
I work on google maps. You can fix the geocode of your address from google maps itself. If the problem in your town is more systemic than "maps can't locate my road" or "maps misplaces my address on the road", please send me an email with some details and I'll investigate (but I don't work all that close to the geocoding portion of the service), rrenaud@google.com.
My main complaint about Google Maps is how it always assumes I'm in America. I live in Tokyo. My default location is Tokyo. I search for "Ginza". I get a sushi restaurant in Alabama.
If you're logged in there should be a "Set default location" or "Change default location" link on the left side, just bellow "Get directions" and "My places".
MapBox (http://mapbox.com/) hosts tiles (http://tiles.mapbox.com/) that you can use with much less restrictions, and download full tilesets if you want to really hammer a server. There's also a tool (TileMill) for designing maps from scratch or with OpenStreetMap data. It's a self-plug (I work there), but pretty relevant today.
Do wish that they had released the Maps API to the API console in parallel with this announcement so that we'd know what we were in for (or at least given a date for its release so that we know when to check back?). My back of the envelope estimate for our usage is just a little bit terrifying right now.
Also not a fan of having to go through a Sales rep to find out anything about the pricing structure for Premiere.
A storm cloud in the distance but I wouldn't start freaking out just yet. "We will also provide at least 30 days notice on this blog before enforcement of the usage limits and billing for excess usage begins." I hope they'll need to pull a "Qwickster" on this one.
Doesn't really need a blog to translate it.