Google Maps' Moat Is Evaporating(joemorrison.substack.com)
joemorrison.substack.com
Google Maps' Moat Is Evaporating
https://joemorrison.substack.com/p/google-maps-moat-is-evaporating
427 comments
I really want to use Apple Maps full time on my iPhone, just because of how much clutter Google Maps has accumulated on the screen.
If anything tells me Google is losing the ability to innovate it's the newly added pill buttons for pre-set searches and branded POIs on the map. I remember when this was added to MapQuest circa 2010, and was the transition from trying to have an excellent mapping experience to just adding revenue focused features.
Google is still operating from an extremely strong position, by nature of having nearly infinite budget and owning their mapping and search dataset. It's concerning that the iOS app seems to be suffering from the same issues that preceded the gradual downfall of MapQuest from a product perspective.
If anything tells me Google is losing the ability to innovate it's the newly added pill buttons for pre-set searches and branded POIs on the map. I remember when this was added to MapQuest circa 2010, and was the transition from trying to have an excellent mapping experience to just adding revenue focused features.
Google is still operating from an extremely strong position, by nature of having nearly infinite budget and owning their mapping and search dataset. It's concerning that the iOS app seems to be suffering from the same issues that preceded the gradual downfall of MapQuest from a product perspective.
Great article, but I think the conclusion about Google Maps languishing because it’s a low margin product is wrong.
While Google Maps as a distinct product has a bottom and top line I think the author underestimates the value that the underlying places data has for Google’s core search product.
The local place data is _massivley_ important to search. I don’t see any foreseeable way Google let’s the product that generates it languish. It’s far too strategic for them. Not to mention the amount of brand value maps has for them as an ecosystem.
My view is that as Google’s Maps advantage dwindles, they’ll do more and more strategic acquisitions of places data sources. Tying more data into their closed ecosystem. Whether the new tech anti-trust landscape in the US and EU allows this is another question.
While Google Maps as a distinct product has a bottom and top line I think the author underestimates the value that the underlying places data has for Google’s core search product.
The local place data is _massivley_ important to search. I don’t see any foreseeable way Google let’s the product that generates it languish. It’s far too strategic for them. Not to mention the amount of brand value maps has for them as an ecosystem.
My view is that as Google’s Maps advantage dwindles, they’ll do more and more strategic acquisitions of places data sources. Tying more data into their closed ecosystem. Whether the new tech anti-trust landscape in the US and EU allows this is another question.
What’s up HN?! Always fun to find an article of mine on here. What I’ve learned so far:
I’ve underestimated the important of their place data to the rest of the search apparatus, and likely underestimated the relevance of Maps to their autonomous vehicle strategy.
What I remain curious about that I’d love your speculation on: why do they bother with the Google Maps Platform at all? They have plenty of ways to monetize with ad targeting, so why go through the headache of opening it up to other businesses?
I’ve underestimated the important of their place data to the rest of the search apparatus, and likely underestimated the relevance of Maps to their autonomous vehicle strategy.
What I remain curious about that I’d love your speculation on: why do they bother with the Google Maps Platform at all? They have plenty of ways to monetize with ad targeting, so why go through the headache of opening it up to other businesses?
> Meanwhile at an office down the hall, Azure Maps has been humming along quietly since 2018, beating AWS to the punch by over two years.
Microsoft's actually been in the geo API game for significantly longer, pre-Google Maps. MapPoint started as a desktop GIS product in 2000 that let businesses create desktop maps, and they've had an online API since 2002... it started as a SOAP API: https://news.microsoft.com/2002/04/10/microsoft-announces-ma...
But it was pre-AJAX so Google Maps stole much of their mindshare. And it's been renamed so many times that no one can keep track of it (MapPoint, Bing Maps/Live Maps/MSN Maps, Azure Maps). Though the pricing and documentation has consistently been reasonable.
Microsoft's actually been in the geo API game for significantly longer, pre-Google Maps. MapPoint started as a desktop GIS product in 2000 that let businesses create desktop maps, and they've had an online API since 2002... it started as a SOAP API: https://news.microsoft.com/2002/04/10/microsoft-announces-ma...
But it was pre-AJAX so Google Maps stole much of their mindshare. And it's been renamed so many times that no one can keep track of it (MapPoint, Bing Maps/Live Maps/MSN Maps, Azure Maps). Though the pricing and documentation has consistently been reasonable.
I live in London and for commuting around via public transport OR walking I always use Citymapper, which is massively superior than Google Maps for directions. I suspect they are using Googles APIs underneath it all, but their UI is better than Google maps and also their recommended public transport directions are orders of magnitude better than Google Maps.
Citymapper will get you between any two locations in London and give you many options based on cost or time or preference for one transport type (e.g. preferring tubes versus buses). Citymapper also gives multi-modal directions: e.g. walking PLUS take an Uber. Or walking, tube, then a cab. Not only this, but it ALSO tells you where you should sit on a train so you have the least walk when you exit the train.
Citymapper will get you between any two locations in London and give you many options based on cost or time or preference for one transport type (e.g. preferring tubes versus buses). Citymapper also gives multi-modal directions: e.g. walking PLUS take an Uber. Or walking, tube, then a cab. Not only this, but it ALSO tells you where you should sit on a train so you have the least walk when you exit the train.
Apple is so far behind outside of the places they prioritize that it'll take decades at this pace to get on Google Maps' current level.
It's really useless in places like Eastern Europe, where Apple has no interest in supporting and isn't likely that it'll turn that around any time soon.
It's really useless in places like Eastern Europe, where Apple has no interest in supporting and isn't likely that it'll turn that around any time soon.
> They appear, to me, to be acting from a place of fear and conservatism rather than innovation.
Bingo. Here's a fun exercise: try to find the last genuinely useful feature added to the Maps API in the release notes.
https://developers.google.com/maps/documentation/javascript/...
To give you a taste, here's the largest change made in 3.42:
> InfoWindows now have a default max width of 648px which can be overridden by setting the InfoWindow maxWidth property. The width of an InfoWindow can now exceed 648px but will still be limited by the width of the map. (Previously, Infowindows were always restricted to the lesser of 648px or the map width.)
Also note that the minor version (0.x) is incremented quarterly, and we're up to 3.43 now, so over 10 years.
Bingo. Here's a fun exercise: try to find the last genuinely useful feature added to the Maps API in the release notes.
https://developers.google.com/maps/documentation/javascript/...
To give you a taste, here's the largest change made in 3.42:
> InfoWindows now have a default max width of 648px which can be overridden by setting the InfoWindow maxWidth property. The width of an InfoWindow can now exceed 648px but will still be limited by the width of the map. (Previously, Infowindows were always restricted to the lesser of 648px or the map width.)
Also note that the minor version (0.x) is incremented quarterly, and we're up to 3.43 now, so over 10 years.
These street maps are becoming insanely precise, probably helped by the competition. I see that Google Maps is now testing "detailed street maps" in some cities: https://www.justinobeirne.com/google-testing-detailed-street...
Just a thought: maybe they are scanning cities for their self-driving cars and are also using this data for Maps?
Just a thought: maybe they are scanning cities for their self-driving cars and are also using this data for Maps?
Here's a little story: I run a tiny digital agency. Our clients are all non technical. They all want maps of some form. Until Google did their infamous pricing thing, all was well: we could set up as many nice embedded maps on websites as needed, clients all happy.
Then the price / key requirement thing happened. Turns out you have to navigate through the absolute hell hole that is Google Cloud Console or whatever they call it this week. It involves permissions, and users and keys and all manner of layers that only a hardcore Dev can cope with. Google don't let us do it on client's behalf - I certainly can't ask client to do it for themselves. Stalemate.
So, yeh, MapBox here we come.
Then the price / key requirement thing happened. Turns out you have to navigate through the absolute hell hole that is Google Cloud Console or whatever they call it this week. It involves permissions, and users and keys and all manner of layers that only a hardcore Dev can cope with. Google don't let us do it on client's behalf - I certainly can't ask client to do it for themselves. Stalemate.
So, yeh, MapBox here we come.
This is a familiar refrain I've recently begun to hear about many of Google's products and I'm starting to see the point. Technology has evolved significantly in the last ten years. It is now significantly easier to build a search engine with its own index and crawl the entirety of the Internet than it was ten years ago. Maps too, will go that way soon. It took IBM and Intel, who were regarded the same way Google is now, half a century to be dethroned. For Google, this might happen on a more accelerated schedule. I suspect there will be a swarm of startups biting at its heels within the next five years.
I tried opening Apple Maps, and they are still way behind Google in almost everything where I live in (EU, Italy). For instance, Apple Maps correctly pins my parent's house but they show it as being several hundred metres away from the street (it's not). Look around also isn't available in my area, and as far as I could see it's not even available in Milan (there's 3D available, though). The only areas I saw the binoculars icon turning on was while hovering over London and Dublin, which puts it firmly at least 10 years behind what Google's been offering in Europe, which has close to complete coverage of rural areas, including in certain parts even roads not fully paved.
I think that outside the USA (and China, where Google maps isn't available) it will be a while before people will actually consider switching. The only thing I'm sure of is that I haven't seen any iPhone user using Apple Maps in my area since, I don't know, ever.
I think that outside the USA (and China, where Google maps isn't available) it will be a while before people will actually consider switching. The only thing I'm sure of is that I haven't seen any iPhone user using Apple Maps in my area since, I don't know, ever.
I've avoided using Google Maps API for Apple devices.
The main reason, was that I knew someone that used it, and got the "Google $urpri$e," where, all of a sudden, they owe four-to-five figures a month, for a narrow-demographic $0.99 app. They switched over to MapBox. MapBox doesn't have quite the feature set and quality that GM does, but it's a great library, and won't give you sticker shock.
Google has a lot of money for a reason.
I have found that Apple Maps is fine for my purposes. I don't really need some of the fancier stuff that GM and MapBox gives you. If I needed the fancier stuff, I'd probably go with MapBox.
The main reason, was that I knew someone that used it, and got the "Google $urpri$e," where, all of a sudden, they owe four-to-five figures a month, for a narrow-demographic $0.99 app. They switched over to MapBox. MapBox doesn't have quite the feature set and quality that GM does, but it's a great library, and won't give you sticker shock.
Google has a lot of money for a reason.
I have found that Apple Maps is fine for my purposes. I don't really need some of the fancier stuff that GM and MapBox gives you. If I needed the fancier stuff, I'd probably go with MapBox.
Meta: could someone explain/clarify the sentence "Back in 2012, I had not yet been born (I am, in fact, only six years old)." please? I highly doubt a 6-years-old wrote this article, so is it a joke? Or is it alluding to some mapping alternative being 6 years old?
Personally, what I find mind-boggling is the inability of Google to capitalize on Map's potential.
If improved, it could concurrence at least TripAdvisor and Yelp easily (and maybe Booking too, if they were willing to invest in customer service - which I doubt). Maps has the potential to be integrated (or better integrated) everywhere in Google's digital estate (search, GMail, Android, even YouTube).
Simarly, simplifying the creation, sharing, and most of all, discovery of user-made maps could boost it enormously.
Instead, it languishes. The UX is incredibly frustrating. Just sharing a location on a map can be non-trivial.
Making it more difficult to build on top of maps was an incredibly bone-headed move as well.
If improved, it could concurrence at least TripAdvisor and Yelp easily (and maybe Booking too, if they were willing to invest in customer service - which I doubt). Maps has the potential to be integrated (or better integrated) everywhere in Google's digital estate (search, GMail, Android, even YouTube).
Simarly, simplifying the creation, sharing, and most of all, discovery of user-made maps could boost it enormously.
Instead, it languishes. The UX is incredibly frustrating. Just sharing a location on a map can be non-trivial.
Making it more difficult to build on top of maps was an incredibly bone-headed move as well.
Related... when I'm in the car, and not, I often use the google assistant to search... for at least a couple months, it's gotten to be almost worse than useless as the bottom 1/3 is covered by google buttons over the stuff I'm searching for, and the mini map view at the top doesn't easily expand to the screen for pan/zoom.
The assistant should be much easier to use, and while not the best, I admit I have tried to use it driving, and now I pretty much have to pull over and use it, because "OK Google give me directions to X on Y st in City" no longer does what it should.
The assistant should be much easier to use, and while not the best, I admit I have tried to use it driving, and now I pretty much have to pull over and use it, because "OK Google give me directions to X on Y st in City" no longer does what it should.
Some clarification needs to be mentioned. Also, this is an issue with the industry. There are more companies out there than MFANG. Our industry should acknowledge other players out there.
Bing Maps and Facebook uses HERE maps under the covers. That's a company that has had the people to really knock mapping out of the park. However, you have some major mismanagement there and lack luster direction that made google maps continue to exist. (They have lead the way in offline mapping.. no joke in 2010 you could install it on your phone, fly from Europe to the US and drive to Canada and back without having data or having to micromanage your "offline data storage") Also, to add insult to injury.. you used to get here.com to point to your customer facing maps, now you have to go "wego.here.com" if you even want to use the consumer maps anymore.
Bing Maps and Facebook uses HERE maps under the covers. That's a company that has had the people to really knock mapping out of the park. However, you have some major mismanagement there and lack luster direction that made google maps continue to exist. (They have lead the way in offline mapping.. no joke in 2010 you could install it on your phone, fly from Europe to the US and drive to Canada and back without having data or having to micromanage your "offline data storage") Also, to add insult to injury.. you used to get here.com to point to your customer facing maps, now you have to go "wego.here.com" if you even want to use the consumer maps anymore.
A recent, complementary piece by the same author: OpenStreetMap is Having a Moment
https://news.ycombinator.com/item?id=25137235
https://news.ycombinator.com/item?id=25137235
Apple Maps is for from parity with Google Maps. I use both extensively. I tend to use Apple Maps more since I like how the Apple Watch integration works, and prefer the UI on CarPlay, among some other benefits. Though I usually always and going to places I know that are not far, and use it more so I now if say their is unexpected traffic and I should take a different route, and well as gentle reminder that a turn is coming up when the kids are distracting me. I'm in New England we have the 'new' maps, though when going some where new I more often than not still always run into issues with Apple Maps, where the place I'm going the location on the map isn't correct, or directions stop prematurely when there is a small road your suppsoe to take or who knows what else, where is I almost never have those issues with Google Maps/Waze. Also anytime say I store closes and moves, or a new store opens, Google Maps shows it within a day or so, where is Apple Maps most often can go weeks or months, and often I am the one submitting the issue to Apple Maps (Which they then correct but takes a while vs Google Maps where I submit an issue and its fixed very quickly). So while Apple Maps is getting much better, there reporting feature is far behind Google Maps and until that changes I don't see how they will be able to be as good. And for many of us Waze is preferred over Google Maps or Apple Maps, when on the highway, since knowing when police are ahead, or a road hazard in real time is invaluable. Though for me, like many Apple Maps is usually good enough, and when it shows the wrong end result, I simply quickly open up Waze or Google Maps to get me the last mile.
One thing I have noticed and found really irritating is the degradation of the street view functionality in Google maps. They have got lazy. I used to be able to click or keyboard navigate through a route but now it either stops responding to clicks, transports me onto a nearby motorway randomly,or gets massively unresponsive.
I am very glad there is some competition coming, as good as Google maps is they are clearly resting on their laurels and not bothering to fix blatant issues.
Competition is good.
I am very glad there is some competition coming, as good as Google maps is they are clearly resting on their laurels and not bothering to fix blatant issues.
Competition is good.
I went to look at apple maps
https://www.apple.com/maps/
All I see is an advert page telling me "You’ll like where this is going" and "get much easier"
There's a tiny "open maps" in the top right corner which links to https://maps.apple.com/go, but that just reloads the advert page
OSM and Google Maps work just fine
https://www.apple.com/maps/
All I see is an advert page telling me "You’ll like where this is going" and "get much easier"
There's a tiny "open maps" in the top right corner which links to https://maps.apple.com/go, but that just reloads the advert page
OSM and Google Maps work just fine
We spend enough on Google Maps API that I've been looking at alternative. We're more interested in fuzzy searching, addresses and points of interest.
Street numbers seems like a pretty big weakness for open source datasets. If all you have is the name of a 1km street, you only get 1km precision.
The APIs from other commercial providers (e.g. TomTom) have better data, but noticeably worse search. I don't think they make their raw data available except for very large deals.
Street numbers seems like a pretty big weakness for open source datasets. If all you have is the name of a 1km street, you only get 1km precision.
The APIs from other commercial providers (e.g. TomTom) have better data, but noticeably worse search. I don't think they make their raw data available except for very large deals.
Isn't the whole moat of google maps that it's part of google? when I search I don't want to have to use different websites for each type of search, right now google is the leader but even if they become only mediocre at maps but stay at no. 1 at search they're very safe.
I don't see any future where google creates a special modal box for apple maps on it's search results page no matter who is deemed "the best map"
I don't see any future where google creates a special modal box for apple maps on it's search results page no matter who is deemed "the best map"
Can't think of a product they've fucked up more with every iteration than Maps. Could argue Search here but Maps has been absurd.
It started around the time they were pushing Material design and hasn't stopped since then.
It started around the time they were pushing Material design and hasn't stopped since then.
> You know Google Maps? What I do is, like, build little pieces of Google Maps over and over for people who need them but can’t just use Google Maps because they’re not allowed to for some reason, or another.
I've been working with web mapping for 15 years and I've never thought of saying it because it's never been true for hundreds of applications. Google Maps has always been the worst map client to integrate with. OpenLayers, leaflet, MapBox and myriads of alternative data sources, layers, behaviors and widgets are more suitable for custom applications.
Some people seem to love a particular Google layer, like parcels or businesses, but it never seemed vital unless they were trying replicate some Google Maps experience, in which case I would just advise to send a route or street view request directly to the Google Maps website or app, and be done with it.
The latest fad seems to be 3D terrain and 3D buildings. It's so amazing in the product demso, but the real world usefulness?
I've been working with web mapping for 15 years and I've never thought of saying it because it's never been true for hundreds of applications. Google Maps has always been the worst map client to integrate with. OpenLayers, leaflet, MapBox and myriads of alternative data sources, layers, behaviors and widgets are more suitable for custom applications.
Some people seem to love a particular Google layer, like parcels or businesses, but it never seemed vital unless they were trying replicate some Google Maps experience, in which case I would just advise to send a route or street view request directly to the Google Maps website or app, and be done with it.
The latest fad seems to be 3D terrain and 3D buildings. It's so amazing in the product demso, but the real world usefulness?
I would switch to Apple Maps if they adjusted your estimated position while driving. When you're following directions on Google Maps, the app seems to estimate your location based on your current speed and the road you're on. Apple Maps doesn't do this (or doesn't do it as much) so it causes me to miss turns.
The more competition the better, though article’s take might be somewhat rosy-eyed.
In the end-user apps domain, alternatives including Apple Maps lose to Google Maps miserably in all regions I have been to in the past 2–3 years with the exception of mainland China. If I need to get somewhere and accurately estimate arrival time, I am not happy feeding Google my movement data but I will since it’ll be the most reliable option available no matter where I am.
As far as developer-oriented solutions, I know that Algolia Places shut down earlier this year citing insufficient quality of OSM data[0]. If Mapbox is indeed getting there and we’re about to have some proper competition in this space, I’m all for it.
[0] https://www.algolia.com/blog/sunseting-our-places-feature/
In the end-user apps domain, alternatives including Apple Maps lose to Google Maps miserably in all regions I have been to in the past 2–3 years with the exception of mainland China. If I need to get somewhere and accurately estimate arrival time, I am not happy feeding Google my movement data but I will since it’ll be the most reliable option available no matter where I am.
As far as developer-oriented solutions, I know that Algolia Places shut down earlier this year citing insufficient quality of OSM data[0]. If Mapbox is indeed getting there and we’re about to have some proper competition in this space, I’m all for it.
[0] https://www.algolia.com/blog/sunseting-our-places-feature/
I traveled to Beijing and Shanghai and Google maps was effectively worthless. Apple Maps on the other hand was accurate and included accurate mass transit directions.
To me, the killer maps app would be the ability to type a consumer product into the maps search bar and have it display on the map all businesses (likely) selling that product. There are still lots of items that are not on Amazon or that I cannot wait two days for.
Every time I try this now, I'm amazed by how bad the results are. First, I think that the simplest ML tools would be able to (say) identify which stores are grocery stores by their name, yet Google Maps is unreliable at this. Second, merchants have a big incentive to tell Google what they sell (either typically or, more ambitiously, with an up-to-date inventory API), if only Google would give them a portal to do so.
Every time I try this now, I'm amazed by how bad the results are. First, I think that the simplest ML tools would be able to (say) identify which stores are grocery stores by their name, yet Google Maps is unreliable at this. Second, merchants have a big incentive to tell Google what they sell (either typically or, more ambitiously, with an up-to-date inventory API), if only Google would give them a portal to do so.
Only tangentially related, but has anyone noticed that the Google Maps app for iOS has become significantly worse in the last few years? They have crammed so many touchpoints on to the map that it is impossible to touch anything of value - I can't tell you how many times I have somehow managed to select an entire county (something I don't think I've ever intentionally done) on my iPhone instead of a POI, which then zooms out to show me the whole county and removes all of the POIs I was looking at in the first place!
Is this a problem for others as well?
Is this a problem for others as well?
Google's choice to bump these prices made me realize that they could do the same thing with their compute cloud pricing or other APIs, again with little notice. That, in turn, led to the realization that we needed to remove our reliance on Google's entire stack, and quickly, or we'd be in a terrible bargaining position later.
This one choice by the maps team likely cost Google significantly in the medium to long term, if any other companies realized the same things we did.