I worked on Google Maps monetization, and then on Maps itself.
Monetization was a dismal failure. I don't know how well they're doing now, but Maps was a gigantic money-loser, forever. I'd be a little surprised if it didn't still lose money, but maybe less. I don't what those "pin ads" cost, but I'd bet it's way less than a search ad.
If you don't believe that, that's fine. "What about indirect revenue?" you ask? Google consciously does not want to estimate that, because such a document could be discovered in patent litigation. As it is, there are tons of patent lawsuits about Maps, and the damage claims always tried to get at Ads revenue, because Maps revenue was nil.
Caveat: I could be way out of date here. I've been retired a while now.
As for the UX: "enshittification" and big-company bureaucracy describe it pretty well.
I know, because they tried to sell it to me when I was at a megacorp, they resell the real time location and local data correlated to user characteristics, as data service to large discrete customers. This includes constant real time information about where people are traveling to, businesses being frequented by demographics, with a sophisticated query interface for asking relatively complex questions of maps derived data. It has some differential privacy layers but we suspected it could be attacked to suss out individuals with sufficient care and effort. Our impression was “holy shit that’s creepy” and it convinced us to never put our sensitive workloads or data in any Google product, cloud or other.
Not directly related but last time I was this creeped out is when a antivirus company wanted to sell us crazy amount of data on our customers.
Apparently AV companies lose money on the customer endpoint protection (it’s basically for free!) but make it up on selling users data. Also on the business licenses though
GeForce experience tracks every single window you have open, sending the title and window information, along with every single spot that you click on your screen to Nvidia, tied to your now mandatory account.
That was a couple years back. Who knows how much more invasive it’s gotten since.
Oh. And this data is considered to be “valuable debugging information”, so is not prevented by opting out of analytics.
There is absolutely no benefit to me the user to force me to create an account. What a deeply frustrating user experience (especially when I have to dig around for my password just to update a graphics driver?!) I hope someone wakes up at Nvidia and removes the account requirement. Why do they even do it?
The one thing that I like about mandatory accounts like this is that they're a strong indication that the company is abusive and you should remove any of their software immediately. It makes the problem obvious.
Nvidia is on top of the world right now and companies on the rise often make really stupid long-term customer alienating decisions. Nvidia won't miss any on of as an individual customer at all. For these types of situations, usually at some point the things that were pushing the company upwards slow or stop while the damage from their arrogance accumulates enough to start causing real harm. After some foot dragging and false reforms, the company eventually has to do something to repair the damage and by then, those who enacted those policies are usually gone.
Good to know! Uninstalling when I get home. Windows has really become a big bag of certified malware and adware. I always feel a bit dirty when I play a game on my windows pc, and now I know why.
The irony here is that you will get NVIDIA (and AMD) drivers vendored through Windows Update automatically by default. These work fine for 3D acceleration, and don't install GFE or require an account.
You'll just be a few versions behind, not sure what their release cadence is for this. But as far as I know this is a net positive to the Windows PC user and potentially eliminates needing to deal with GFE/Adrenalin as long as you are satisfied with the builds Microsoft is shipping you.
>Windows has really become a big bag of certified malware and adware.
This 100%. I would include Azure too.
I actively try not to use Windows every time I get a chance. I am so glad a company like Apple exists. I agree Apple might not be 100% right on privacy but they are pretty darn close that I a live with.
How I was wrong is was in the “valuable debugging information”. Because you can actually disable the sending of debug and crash statistics, but you cannot disable the collection of tracking everything you do except for password boxes!
The original source on the telemetry was a print magazine, CanardPC, but this is referenced quite a bit.
I was additionally wrong the GeForce Experience matters. This is currently untrue. After the original article was released of how much spying Nvidia does, the telemetry package was moved to the driver rather than GeForce experience.
Long story short: if you have an Nvidia video card and have not actively disabled nvtmmon then Nvidia is tracking every window you open, how long it’s open, which windows have focus and for how long, window locations titles etc, along with clicks (but not keyboard strokes) etc. *It is not only tracking games. It tracks everything. It does this with just the driver*. If you have GeForce experience, then that information is tied to your sensitive personal information. If you do not opt out, they sell that to advertisers. If you do opt out, the collection happens anyway, but they claim they’re not sending it to third parties.
Where does your link support what you are claiming? Clicking "see more" on all the sections and searching, there is no mention of "valuable debubbing information". Under "Using NVIDIA Products" section it lists GeForce experience, but not plain graphics cards or their drivers. I have no nvtmmon.exe running and the folder
I’ve just learned: you don’t. I thought this was limited to just GeForce experience, but it’s not.
You need to kill Nvidia telemetry services (one site has it listed as nmtvmon).
Another option would be to block traffic to Nvidia from leaving your network (pihole captures at least some Nvidia telemetry, not sure if it blocks all). I don’t have a list of urls or IPs to look out for.
I use glasswire firewall and I just checked nvidia's activity and it doesn't even amount to a megabyte (uploaded) in the last week. I'm thinking this is tinfoil hat stuff.
Assuming they're using halfway decent cryptography, then doing this is nontrivial. But there are probably people working on it, and if so, they may eventually succeed.
Technically I don’t think it’s resold, as google hasn’t bought the data from anyone. But I think what the author meant is that Google also uses the data for ads, etc and then sells it.
This seems wrong. I don't think the average business owner has the capability to reverse engineer the anonymous tokens that represent hash IDs. Also, it's good to go in and reset that occasionally with Google. I gave up on it and switched to apple, tho. Far as I can tell, they don't ever sell that info, even in anonymous hash # form. I'm happy to be proven wrong other than "wHy wOuLd yOu tRuSt aPPle?" .
However it wasn’t clear how high quality the differential privacy algorithm was or if it was attackable. But the nefariousness of its use against groups and demographics was fairly apparent.
It seems like the person you're responding to has a problem with them selling their data in real-time to 3rd parties, rather than just any Google employee knowing where they are heading.
Yes that’s the specific issue. We didn’t fear specifically they could read our data; the key management systems available appeared sufficient at some level (although the gcp audit system had some issues, particularly access audit is available as a IAM decision log rather than point of use, so any access bypassing IAM or that failed in the service but succeeded in an IAM call would appear a positive access, etc - aws does the right thing here). More of concern is they appeared more than happy to harvest utilization information for the profit of others, and in our business that was a potentially serious side channel.
Surely they can. Who would be locking them out? Another google employee. So the useful questions are which google employees can look up people's locations, and through what process?
Some people need to be able to do that as part of their job. Either directly for some good reason, or indirectly (they are a DB admin). The question is what controls they have to ensure only those people can look such things up and that they don't abuse those rights.
that's completely different. Can an accountant or marketing person look up a persons location vs can any developer vs can any map developer vs specific people with proper access controls
From what I've heard that type of information is only held on a special higher-security "logs" cluster and the code accessing that data is subject to additional review by Google's privacy division before it is allowed access to the data. I think there may be special ways to manually access some of that data, but even requesting that capability would automatically trigger an audit after the fact.
Google Maps design is becoming less detailed over the years. It's really apparent when compared to OSM.
It is necessary to search for a lot of things in Google maps as they are otherwise not visible.
You can't really navigate off Google maps without turn-by-turn navigation.
I wonder whether this is a strategic drive from Google to make maps more like search to enable advertising or whether they truly think this is the superior user interface.
Yeah it's weird, very often a venue on a street is simply not visible unless zoomed in TO THE MAX. And then it's not any weird obscure thing, it's a restaurant or coffeeshop that's been there for years.
What's even worse is that often I'll click "restaurants", zoom in all the way, and then click "Search this area", yet Google will still refuse to show some popular restaurant, even when there are no other restaurants displayed in the window.
The only way to get those restaurants to show up is to search them by name.
Not sure if they're intentionally penalizing certain restaurants, but it's pretty bad.
Been noticing this, myself. E.g. for a local used bookstore that's reasonably popular and has been there for decades. I know the owner and can guess -- though I haven't asked -- that they aren't giving Google much if any money.
So, no Maps presence for you! (Except via direct search within Maps.) Just my anecdotal experience, but one of several similar on Maps that leave me thinking they "have that feel" to them.
Also the "cuisine" filters for restaurants appears to be completely irrelevant. You will still get results for Chinese restaurants even though you have selected "Pizza".
It would seem that the revenue-optimizing ML has learned that the median user could really go for Chinese take-out right about now (for all values of now), no matter what sort of food they started out looking for, so it's always going to show up.
I know median and mediocrity already share a common root, but I wish there were a clever portmanteau describing ML-optimizing for the median user leading to a mediocre experience. "Medianocre" just sounds like an eggcorn, so I presume 99% of readers would take it as a mistake instead of a neologism.
FWIW Apple does the same thing to me in my neighborhood. I live in Brooklyn where there are restaurants on every block but it'll zoom out and show ~15 restaurants scattered over a mile with half in Manhattan. Pretty unusable.
I love OSM and use it all the time for hiking, but even though their restaurant and hotel listings have good coverage, they're sadly not much use to me without the reviews.
Short of scraping reviews from Google and importing them into OSM for the area in which I'm traveling, I'm not sure there is a good short-term solution.
I get much better results by clicking on the text input field and pressing enter every time after moving/zooming instead of clicking on "search this area".
There was some articles on here about how maps have been changing over the years (also other map apps). The gist is that why would you need to show street names anymore? On printed maps that's the only thing you have, but that's not how most people navigate anymore. They search for a place or POIs and then use navigation.
So the things most people care about have changed, streets are less important now and things like attractions or "active areas" are more important. Yeah, some people still read maps the "old way", but now probably those are in the minority so catering to them isn't worth it anymore.
Of course that's only one aspect next to the question of what makes the most ad money.
I've sometimes noticed friends navigating (on foot) in a residential area staring at their phone until the blue dot is on the turn they need to make.
I would usually look at the map briefly, see "we need Farm Lane, should be 4th on the right" and then look at the street name signs. In a non-residential area I'll often do this on foot, but if I'm cycling I'm more likely to think "Right after the paint shop" or similar.
Perhaps not a large use case, but I use street names regularly when discussing bicycle routes. "How to get from Town A to Town B by bicycle" often relies on street names - take Rt 123 for 2 miles, turn left on Fancy Farm Lane, etc. The route-finding often keeps you on Rt 123 for the whole route, but it's a busy dual carriageway and really not suitable for bikes.
I think a big problem is that Maps' bicycle routing would prefer to send you down a 55mph 4-lane street with a "sharrow" in the right lane, because that's a "bike friendly street", instead of the 25mph quiet residential street that parallels it.
There were always people who were like that it's just that it was harder with paper maps, at least that's my view.
For those old enough, just think back to all the times your "navigator" wouldn't just tell you the street name or even "turn at the third intersection after this one " and instead waited until just moments before you reached it and said "Here" like a jack hole.
> The gist is that why would you need to show street names anymore?
I need the street names. They're authoritative. Landmarks (or, worse, businesses used as landmarks) are things that are occasionally useful to me in the right circumstance, but they're too dynamic to really rely on.
If a map application doesn't make it easy to know the street names, then it's one that isn't useful to me.
If you haven't read 1984, this is really confusing, but a "memory hole" is a hole into which you place something you don't want to have any memory of, destroying it forever. The ironic name is an example of "doublethink."
And those who've actually read it also know that Newspeak is also a term he coined and that since we don't actually
have newspeak, the natural extrapolation of the term Doublethink into a spoken application is the term Doublespeak....
Right?
On the flipside digital maps can progressively show layers on user demand. You can pull up the OSM map with all the street names. If you've got a bit of patience to learn tools like QGIS (or ArcGIS) you have more power than ever to build personal maps that work best for how you personally navigate spaces beyond the lowest-common denominator that used to be print maps.
It's maybe a bit of a shame that the raw tools (like OSM, QGIS, ArcGIS, even what is left of Google Earth) aren't easy to learn and equally accessible to everyone and the incentives of the big maps (Google Maps, Apple Maps, etc) is its own drive to a different lowest common denominator, but digital maps have such tools in the first place. (It was far more expensive to get custom printed maps. AAA triptychs are interesting thing to compare here.)
You need street names as I see the name on a map on iPhone plan in my head a route and see the street name so as to know where I turn - or how far you are along a road. GPS is not that accurate.
Yea it's got some really odd behaviors that I can only attribute to Yelp style behavior from Google.
I was literally on the same block as a restaurant and it wouldn't show in maps at all, though the other 3 restaurants on the block were, and the only way to get it to show up was to Google it then click on the maps link. As soon as I took focus off the restaurant it disappeared again.
I can only presume it's because they hadn't paid Google or were being punished for some reason. It lead to me looking for more examples and found many places, including some major ones in the city I'm visiting, were excluded for no apparent reason, even ones with 4.9 ratings, and the only correlation between them was that they either were inherently more "indy" than the rest and thus likely to avoid engaging in certain aspects of Google or were close to a more powerful competitor.
This was particularly shocking when Wendy's wasn't being shown at a strip mall but the Burger King was....
I'm not sure why they could add "personalities" or "modes" like simple, tourist, night-life, etc. Seems rather obvious to just about anyone who thinks more than 30 seconds about it
I was working on web maps before google maps steamrolled (effectively killed) the competition. The big thing for us was that we couldn't compete with the storage and licensing requirements for satellite imagery. I suspect it was the same for other local competitors at the time: our local data was objectively better, but google's was just sexier, and you couldn't compete with free.
I've long since switched fields, but the irony is that the company I was working for survived and I heard maps for various business use is now much more expensive than our offer used to be and several large clients eventually switched back.
As a user, I loved maps when it came out. The pre-rendered tiled maps were clean and fast. The web UI was clean and responsive. Streetview was (and still is) absolutely mind boggling. Heck, I even applied to work for google maps at the time.
Nowdays maps is an absolute shitshow. I find it utterly unusable for almost any purpose except for streetview, which is the only reason I still know how it looks like.
Peole forget that millions of people submit data to Google Maps for free. I placed all the local postboxes on google maps, people sibmit information about their businesss, opening hours, etc. This is hugely valuable.
Also google grew laxy. Not tskes them a year to add a royal mail postbox
Compare the experience of adding stuff to OSM and to Google Maps. Much easier.
OSM is for nerds. Very hard to add even basic info. Google Maps it’s trivial.
Same as uploading photo to Google Maps is easier than adding it to WikiMedia Commons. Such is life.
(Ironically one of the biggest sources of photos is commons is Panoramio; before Google merged it with Google Maps, the photos there defaulted to Creative Commons.)
How do I update stuff in Google Maps? For OSM, its a case of opening Street Complete and tapping on things.
To add new features I have to use SCEE, which is more nerdy mainly because it involves using FDroid to install it. But I've no idea how to add, say a park bench to Google Maps.
Just tried it on mobile and that is nice and easy, especially if you want to add a business. Not sure how you add a bench that way as I couldn't see the option, but I might be looking in the wrong place.
You're right though, installing another app store is a non starter.
For sure, but for many people there's only Google Maps, and maybe Apple Maps. On the short term, catering to Google Maps makes sense. They'd add to another project if it were to be popular enough, and the network effect keeps Google Maps popular. Hence they get away with enshittification.
>They'd add to another project if it were to be popular enough, and the network effect keeps Google Maps popular. Hence they get away with enshittification.
Ah yes, the chicken and egg logic of network effect excuses all.
How about if people, idk, just evangelized more about the things they love to be smug about on HN?
People should be shamed off enshittified platforms. Be "that guy" - or be a hypocrite. For Google it might a little more complicated, but for each one of us it's exactly this simple.
OSM has a two-way data sharing agreements directly with (among others) Apple and Bing/Microsoft. Additionally, the "Overture Foundation" with members including AWS/Amazon, Microsoft, Meta, and TomTom have a complicated indirect two-way data sharing agreement with OSM. (Overture Maps are a "curated" subset of OSM data and members are encouraged but not required to make contributions and updates to the maps directly upstream to OSM.)
Google probably sees their data as a moat and likely aren't interested in two-way data sharing. (Given how many of their competitors are now involved directly or indirectly with OSM or Overture, they may be correct.) OSM probably can't legally scrape it without an agreement in place.
Also, everybody by default shares their location when using it, no? So they get live feedback on traffic relatively for free (till some lawsuit will ban it). I can see it often changing and recalculating as I drive, albeit with some lag.
I don't think you can estimate overall worth of Maps to Google directly and easily. Its still #1 driving solution I use, its not perfect in dense traffic in cities but otherwise OK and for free, plus their estimates are pretty accurate.
Agreed. And not just that, but also the fact that people use Google Maps for car navigation means Google can derive the location of traffic jams by comparing current speed with the speed limit.
This is hugely useful for navigation since it means these bottlenecks can be avoided, some thing which is not possible unless this data is available.
If I present you the other side of the coin - we are a small country, where we cant keep up with number of cars of tourists traveling to the next country over summer.
Since the Google Maps is here, not only we have a completely congested highways but also all other roads that were free of tourists before are now blocked. Tourists that are traveling to the next country while their number has doubled as they dont stand in line from border to border but rather destroy any viable traffic for locals.
In some locations it has gone to the point where small villages with historically very narrow roads are literally blocked by **** (pardon my french), using google maps and see that there is a pass (while every road is the same, right) so the police needs to deter tourists to even head there. Wont even go into suffering of locals there.
This is a real issue, but it has nothing to do with Google specifically. You’re basically saying that information on road networks should be obscured or restricted so governments can indirectly control the flow of traffic.
Japan actually has some soft regulations in this space about routing and road visibility (e.g. residential roads are hidden when you’re on thoroughfares). But they’re just suggestions so only domestic companies follow them and Google, Apple, etc still provide their usual global algorithms.
Maybe there’s a good solution to this, but I don’t think it’s getting rid of mapping services and going back to security through obscurity.
No, I am saying that by diverting drivers to regional roads without taking into account local population, this is directly Google fault. Why? Before Google started messing with information about congestion, there was no issue. Regional roads swallowed without any problem those few adventurous tourists that decided to use them. Now you find an average Joe everywhere.
Wanna se another interesting example (use google translate), here are statistics of our volunteer based and taxes paid (for everyone in need, even tourists, including helicopter transport) mountain rescue service: https://www.grzs.si/resevanje/statistika-nesrec/
Check number of accidents since 2006 (google maps came out 2005) and constant rise to 2023 (more than twice the number of interventions). Freaking maps are showing the steepest mountain roads (Alps are not a joke), to parts without phone signal. Google gets you just far enough to get you into troubles, potentially fatal troubles.
> this is directly Google fault.
> Check number of accidents since 2006
I believe it is a funnel of things: Slovenia became part of EU. GPS navigation in cars and smartphones became a thing. Google isn't even the only traffic data provider. Slovenia has RDS-TMC since 2009. So even a boring TomTom will show some form of congestion info.
But I agree with you, that navigation tools should not direct traffic away from the main roads (especial transit roads). Traffic planners should somehow be able to weight streets or routes.
The solution sounds simple. Force non-locals to buy a Swiss-like vignette to drive on those roads. And, make it damn expensive, like 1000 Euros. You can probably put a police checkpoint around a corner about 500 meters after the sign and fine people who are missing the vignette something like 2000 Euros. Finally, use social media and obnoxious free media (City AM and friends) to tell everyone about it. It should take care of itself in about 3 months.
I wonder if it would be possible to get a bunch of old phones and put them by the side of the road with Google maps on, simulating a traffic jam. Or something like that.
Google maps doesn't do this, it's done by lower levels of software, specifically Google Location Service which is built into Google Play Services. You can tinker with maps settings all you want, but the data is coming from the rootkit they bundle on nearly all devices. Many android apps use GLS APIs and won't work without it.
I've been saved from tickets by Google maps several times now, mostly because when it warns of a speed trap it reminds me to look at the speedometer and realize the prevailing traffic speed had been more than 10 km over the limit.
with giant data beast company like google, everything is about indirect revenue. Just because map doesn't make money, doesn't mean it won't drive up other ad revues.
Google uses maps to collect the location data from iOS devices.
So, Apple Maps are useful, because they prevent Google Maps from becoming "too many ads app".
I personally think that ads on maps were a bad idea even for monetization. At glance, this seems to be a plausible ads inventory for local businesses but in reality this doesn't scale very well compared to Search or App ads (or even Display) since most of ads will be limited to a very narrow area so big advertisers don't feel as strongly as other channels. They could design ad campaigns specific to local areas but this increases overall complexity and makes it really hard to plan and optimize its budgets. Also, it's worth mentioning the difficulty of measuring offline conversions accurately... Some degree of automation might help, but at this state it's not really a great place for advertisers.
I hate ads as much as the next guy, but I actually think that if this were better implemented, it wouldn't be that bad.
One way I commonly use google maps is as a restaurant or bar finder. Like "show me open restaurants around me". I even do this when at home and feel like eating out for whatever reason. I don't want to walk too far away or take transit. I just want to see the restaurants in a 10-minute walk radius around where I am which are open and what they serve. Now, sure, I'd like to see all restaurants, not just those which took out ads. But I'd be fine with those being featured at the top of the list.
Instead, what I get, is ads for hotels down the street when I'm already home. Or ads for a carpet store chain, which was closed that particular day (national holiday), which you also can't really miss if you set foot in the square.
Regarding monetization, they could probably figure a way to tell that if I click on the restaurant's pin, then follow directions there and stop at the address, I'm likely dining there.
I do something similar when going around and about in the UK (and to a lesser degree Malta when I happen to be there). I like to visit random places, seeking out locations off the touristy paths. One particular search has helped me over the years: "pubs near me".
Local pubs, even those part of a large chain, tend to give a nice feel of the region and the people who live there. And you can learn interesting details about the local history.
The author notes that Google does seem to be leaning in to the sector where they really make a lot of sense, hotels. I'm going to Delhi for work next week and how I picked a hotel was going to maps, zooming out from the venue a bit, and letting hotel pins appear.
It's also been downright unusable. While navigating it hides all the business names from me so I can't browse the stores are along my way. I constantly have to exit navigation, approximately remember the route, browse the stores, and then re-enter navigation mode. This is only one of many pain points.
Its assistant is also useless.
"OK Google, zoom in the map by a factor of 1.5"
"OK Google, stop hiding the businesses"
"OK Google, add a stop for the last Safeway before the turn onto route 120"
Yes the classic dance of “zoom in slightly further, but this time without arbitrarily removing 3/4’s of my search results that I know exist here, and showing me other random locations”
Or “good lord just show me the street name please”
I’d like to see a defence of the reason why it’s so often a bit of a job to be able to read the street name… on a map. If only for the laugh I would have.
The article complains about 'clutter' - but have you seen a traditional printed map like [1] ? Personally I find them beautiful, but it's a rather Jackson-Pollock-esque beauty.
One of the changes brought about by the digital mapping revolution was to allow much less visually complex maps.
Right but digital map allows you to zoom in and on street level of zoom there is no longer name clutter. There is no need to remove those details to just leave swathes of empty space
Because Google always wants to have its cake and eat it too. Every one of their product tries to be "smart", and it inevitably miserably fails. Google Maps tries to both be a normal map, and then a list of events and places around you, then a shopping guide, then... then it fails at all of these.
Letting you specify on Google Maps a kind of context that would swap between these at will would be an infinitely better experience, but unfortunately it doesn't grant promotions, unlike a blog post about the _brand new smart experience_
They have a blog post about it from way back. Essentially the map changes depending on context so you don't need to see store names if you are zooming down the freeway by the side of it, you might see road names if walking and navigating, but not if you have gps in a car, you might see more restaurants if you came to maps from a restaurant search context etc etc
Without a trace of irony: thank you for explaining to me why I fucking hate Google Maps.
Don't change the context and try to outsmart me without giving me any indication of doing so. Don't attempt to outsmart me without giving any recourse if I don't like it. It only ever leads to information I want to see being hidden. You, Google Maps, do not know what I want better than I fucking well do.
Exactly. Like for example if I'm the passenger and I do want to see the store names along the intended route, and even the ratings and reviews, while the driver is zooming down the freeway to make a decision about where to eat, without taking it off navigation mode.
Oh and not to mention that it sometimes tries to translate street names to my phone's language setting. Leave street names in their original language so I can match it up to the physical signs, goddamnit.
My personal hot take on this is that this is emblematic of present day design strategy: first, abdicate responsibility for meaningful customizability (meaningful = anything but maybe aesthetic theming). “UI is hard.” Once you’ve accepted that it would be impossible to create settings that would be useful and worthwhile, then you just make a bunch of static choices based on what you think the average dummy would want, or write an algorithm to decide it all non-deterministically, which gives a small chance that it will occasionally meet needs more complex than the most basic use case.
You can click the search icon while in navigation mode, to search along the route. It doesn’t work that great as you have to zoom to an aprox area and it seems somewhat arbitrary what comes up. It will show you how many minutes of detour a location would add, which is quite nice when choosing among a bunch
Sometimes I don't know what I'm searching for, I just want to browse what kind of businesses exist in an area along my route to see if it's worth stopping there and taking a break. Also if I use the search feature it auto-zooms out the map and then I have to fight it within hundreds of milliseconds to get back to where I was on the map. Once it auto-zooms fully I may have trouble remembering where exactly along the route I was looking. I may have been searching for food near a particular weirdly-shaped lake along the route that I spent 15 minutes tracing the 500km route at max zoom to find. Once the search auto-zooms back out, my 15 minutes of "work" is instantly wasted and I can't find that lake again.
Well, first world problems for sure. Its free app, you have some extremely specific use-case (even though I have no doubt 100 HN users can post in response that they do exactly same thing, its still fringe overall) and the app doesn't suit that narrow situation that well...
Not surprised, we can come up with probably thousands of similar scenarios when some popular free tool ain't the smoothest to use for 'when I do this but want also that and while doing something else that isn't so well integrated into the flow something happens/doesn't happen but should' situation.
Maybe take a laptop and a mouse next time, 100x more efficient to use. Or use it in browser in tabs, so you can have 20 different locations/views and switch between them.
The search-along-route is more of a 'search-around-my-current position'. If you e.g. need fuel, it lists options for busy towns in the wrong direction rather than the easiest path along your current route. If you need a given RON or niche fuel, you just have to memorise which suppliers sell it & hope that it shows up in the list. The utopian vision that it could figure out the range of the car it's connected to and use this information to decide where would be most efficient to fill up is a pipe dream.
Using Maps while driving for anything other than A-B navigation is generally poor. It seems designed by people who are accustomed to every destination having plentiful dedicated parking. If you want to go to B, then find local parking, you need to invoke dark voodoo.
The worst part is that the Apple & Google duopoly in maps supported by Android Auto and Apple Carplay means there's no way for a superior alternative to succeed in the market. Google even contributed to this by buying Waze. So Maps only needs to be just good enough.
>The worst part is that the Apple & Google duopoly in maps supported by Android Auto and Apple Carplay means there's no way for a superior alternative to succeed in the market.
What does Android Auto and Apple Carplay have to do with it? Aren't both of them available for third party apps to use?
The Apple one is available, theoretically, though you have to specify each time to tell Siri to get directions with something other than Apple, and it also won’t support any features such as the useful “share ETA” which is a real shame. They also thankfully opened up their split-screen mode to third party navigation apps a couple years ago (though, incidentally, that screen is utter trash in my humble opinion because of what it doesn’t include).
I recently watched this show on netflix called "The Billion Dollar Code" that went into the details about maps beginnings and other companies claim to the idea. It was pretty interesting.
I can also recommend "Never Lost Again: The Google Mapping Revolution That Sparked New Industries and Augmented Our Reality" (https://amzn.com/0062673041) about the evolution of Google Earth, Maps, and Street View. Written by one of the founders of Keyhole (the company that built what became Google Earth), the book is mostly about the founding of Keyhole and its mapping technology, but has some insider info about the political maneuverings between Google's Search and Geo teams over owning the map ads because everyone knew it would be a big deal.
Brilliant show, though it is about Google Earth. Made me uninstall Google Earth from my phone though, the picture it paints of Google and their legal people is not pretty.
Ah you're probably right! I know they are separate things but ever since that update ~10 years ago where they brought many Earth features into Maps the product merged into one for me.
If google can't make money with maps, why isn't Maps taken out of google into a charity/non-profit backed thing. There's OpenStreetMap, and Wikipedia wouldn't mind having a Map I guess.
That way it'd help a bit with the monopoly heat and maintaining it could be slightly cheaper. I don't know the legal stuff, but I'd rather have a company supporting a non-profit than burning money internally, so I hope the incentives follow that (yeah, probably not true thanks to lobby and having rich people run governments)
> If google can't make money with maps, why isn't map taken out of google into a charity/non-profit backed thing. There's OpenStreetMap, and Wikipedia wouldn't mind having a Map I guess.
Assuming maps still loses a good chunk of money, how would it survive as a charity or non-profit? Where would the money come from? Who would donate to the charity?
Without donations OpenStreetMap would also lose money. But obviously people (and companies) are much more willing to donate to a project when it's a charity/non-profit, instead of the product of one of the biggest companies in the world.
Maybe people are more willing to donate to a non-profit, but I am not so sure the same goes for companies. Today companies have to pay a license fee if they are using Google maps in their services. A non-profit can of course continue with licensing the service, but if they make that optional, 99% of the companies would stop paying.
So you already can use OpenStreetMap, but don't? Presumable because Google Maps is better? So maybe the answer is that "we" don't make Maps a non-profit majority of people (me and, apparently, you included) actually prefer superior ad-supported product?
This part is probably what keeps Maps valuable as an asset, and it's such a huge moat that letting it go would be irreversible. Building a similar product from scratch takes tremendous capital and local deals, we've seen how much of an effort it is for Apple to even have a viable product, coming up with a competitive offering would require double that.
Even if all counted for Maps was net 0 in the balance sheet, I feel there would be no Google CEO agreeing to let it go just for antitrust relief. They'd probably let Youtube become independant before that.
It takes serious funding to make an excellent mapping app that contains worldwide business info, and a whole hell of a lot of paid review staff. A non-profit/charity can't really tackle this well. And since having a good mapping app is so important to the use of a smartphone generally, it makes sense for the two major smartphone OS vendors to each have a platform-standard mapping app.
Mapping is a data collection/sanitization and content moderation problem. You need a lot of data, much of it user-submitted, and you have a lot of bad actors constantly trying to feed in bad data. I don't see where AI comes in here except maybe for route optimization, which is a tiny percentage of what a mapping app actually does. A mapping app is mostly just a huge database of incredibly useful info that needs to be true (and LLMs cannot help there, they just make shit up).
But is Maps actually a net loss for Google, or are they using Hollywood-style accounting techniques to make it <i>appear</i> unprofitable as an additional disincentive to the lawsuits that AlbertCory mentioned?
It's a massive net loss for Google, where do you think there revenue there would come from? A few API users don't pay for a massive amount of people working on keeping the map data up to date.
Just a couple weeks ago, Google announced they're going to re-photograph Street View in my country, though, which is still running on the initial 2007ish data. How would that be profitable? The idea of virtual ads placed onto pseudo 3D walls was already present when Street View started but apparently didn't deliver economically. Maybe they want to break resistance this time around as they had lots of data protection complaints/requests for blocking a particular lot from displaying unblurred at my place?
Mind that they don't only take pictures. The cars are full of sensors. From GPS to verify coordinates of roads, over Wifi to scan SSIDs for locating devices without GPS to LIDAR to scan 3D features and probably a lot more ...
They already know the photos have the objects they are interested in. Are they still working on motorcycles buses and signs?
The recaptcha part doesn't even depend on recognition, it's just a statistical comparison with what other humans clicked on given the same prompt + some pointer analysis and browser fingerprinting.
Are you talking about Germany? Google never stopped creating new images in Germany, they've just decided they would put the new images online (which they already did).
One of my biggest peeves with Google Maps is its lack of respect for my zoom level. When I tap the centering button, NO, I don't want to zoom in. If I wanted to zoom in, I'd zoom in!
Couple that with the roughly decade-old removal of the +/- zoom buttons, and I basically don't bother with Google Maps any more except for turn-by-turn navigation.
Why do you need +/- zoom buttons? You can just double-tap the screen but hold down the second tap, then slide your finger up or down to zoom in/out as an alternative to pinch-zooming.
speaking of monetization, i’d happily pay a $10 monthly fee for these features as a biker:
- weather along route
Google knows the weather, it knows the route… Just add a layer! better yet suggest a reroute. i pay for a separate app for this currently.
- motorcycle friendly features
bigger buttons for usage with gloves; selection of straight, curvy, unpaved etc roads; right now it’s just Highway or non Highway
- ability to create a group and show their live locations on the map.
Something better than WhatsApp live location. The current approach that Google maps uses is cumbersome and non-intuitive. also if a route is updated, an option to update everyone else’s shared route.
- local info, POI, etc
you know my location and interests, tell me some interesting stuff over Bluetooth about the city or POI’s i’m passing by.
Google play services API's are deliberately designed to be hard to reimplement by third parties. In the case of maps, the placeId of every business or POI is internal to Google, and no other provider can give you a map for a matching placeId.
This isn't unique to Google. For example, the stripe PaymentMethod object is opaque and can't be moved to any other payment provider. That pretty much locks customers into stripe unless you want to force the customer to sign up to your service and enter their billing details again.
It’s an open-source system that is used worldwide, by a few thousand loosely-organized nonprofits.
When Google Maps suddenly required associated credit cards, it caused a great deal of trouble. It still causes problems.
That’s because most of these entities are not 501(c)(3) orgs, with bank accounts and debit/credit cards. In fact, the organization enforces deliberate poverty, amongst its chapters.
Someone needs to add their personal card to the Google Maps API. Since the maintenance people are constantly rotating, this is an issue.
Also, the API has a dashboard like a Space Shuttle cockpit. Most regular schlubs have a difficult time with it.
We use google maps for our non profit event and for the first time this year we had a small charge for our once a year event. We have the same credit card issue even though we’re a 501c3.
Google’s UI is a complete mystery to anyone outside google that doesn’t know the lingo. It seems to keep changing too.
Currently working on multiple apps hitting anywhere between 5k and 150k maps loads a day, costs are zero because we do not do things outside of the free usage tier. The Maps Embed API (the one you see on websites) is pretty clear on that: https://developers.google.com/maps/documentation/embed/usage...
Pricing for the Maps Embed API
All Maps Embed API requests are available at no charge with unlimited usage.
Other usage limits
There are no short-term (queries per second) or long-term (queries per day) limits on the Maps Embed API.
Similar things apply to the Maps SDK for Android. Seeing the watermark means that there's either no API key being sent with the request or, in relatively new developments, that you haven't enabled billing/inputted a credit card. Which is a scary thing to do when you know that you're not doing anything out of the usage limits... but what if Google suddenly decided that ?
From my perspective as an advertiser, mobile/local PPC has taken over, be it Local Service Ads or regular search ads listings that show up in maps.
SMART ads that use your GMB profile instead of your website go GANGBUSTERS for brick and mortar clients, who love them.
I wonder how long you have been gone, and if it's possible that they have taken over so much since you left that your statements here are no longer true.
If it’s not making money anyway it’s strange they pull tricks that are just another thing to make Google look greedy and bad, where they very clearly hide businesses to punish them for not taking out an ad.
It makes Google look like the mob and everyone sees it. Everyone would understand if they would show businesses that pay at the top of the list but somehow instead they went this way, incomprehensible.
> "What about indirect revenue?" you ask? Google consciously does not want to estimate that, because such a document could be discovered in patent litigation.
Could you tell us more how you came to this conclusion? Did someone high up ever literally tell you "we're not doing that because it could show up in discovery"? And if not, why have you concluded that?
Ads and Search use the data from maps to make up the difference.
Even just the ability of search to serve local seeking queries is a huge defensive moat, but the marginal revenue from having a model of the world is worth it.
Given your firsthand experience and the industry's evolving nature, do you think there have been any significant shifts in Google Maps' monetization approach since your time working on it?
Maybe. They have (or have as of 10 years ago when I signed up) a checkmark for I consider this public domain. If you don't check that OSM cannot freely use the data. I only edit OSM so that people can find my house, so I consider it a feature that google can freely use my changes, but not everyone agrees.
OpenStreetMap does hold opening hours of businesses. Here are the details (including opening hours) of my local cafe https://www.openstreetmap.org/way/485163917 I go round entering OSM data using the StreetComplete app. It's slightly awkward if a shop doesn't display their opening hours, because then you have to go in, and I feel I need to buy something and then ask their opening hours.
I do that too (in Germany), and I just… ask them if I may enter their opening hours into OSM. I've never been refused. Sometimes an employee takes an interest and then I take the chance to explain a bit. It's a win-win for both sides, after all—the map users and the shop owners.
In a practical way (instead of just, why not be polite?), the information for permanently displaying to the public may be different than the information relevant to you at a single moment.
If you just ask it without context, "today we'll close at 16:00" is a perfectly reasonable response.
Hopefully a few of them will realize it is easy and so start updating their own information as it changes. Everyone wins - their customers get up to date information and so come when they are open.
Sadly, they don't even attempt to have public transport schedules on the main maps, whereas Google has even added inter-city buses in some places, as well as having an integration with Bolt or other platforms I think.
That, in addition to the lack of business information, business hours, reviews and so on makes OSM a bit inadequate in comparison, which is unfortunate. There have been other maps that build on top of OSM data (which is great), but they're still pretty small in comparison.
It would be interesting to see how OSM would fare if it actually tried to compete with Google: 1st party Android/iOS map with some open dataset for business information, integration with some review platform, aggregated public transport maps with an underlying open dataset and stuffing their pockets as much as possible with money from Bolt/Uber/whatever to ensure longevity of the platform, by adding helpful suggestions for trip routing, as well as aggregating traffic data.
> Sadly, they don't even attempt to have public transport schedules on the main maps, whereas Google has even added inter-city buses in some places
What Google is doing is to integrate public transport data (often in GTFS format). It is not mapped directly by Google, and unlikely that it would be doable in OSM (note that in many cities timetables change weekly, for example as result of construction).
There is some ongoing work to build similar things for OSM-using software, for example Organic Maps is at least trying: https://github.com/organicmaps/organicmaps/issues?q=is%3Aiss... (I bet that pull requests would be welcome, especially if discussed with people developing it)
What kills me is I'd easily pay 50~100 eur a month if Maps had to be non ad subsidized.
This is to me Google's most everyday impacting product, and what makes the difference between me being stranded/lost or making it in time. The stories of people ending in a lake for blindly following directions are cheeky, but decent maps are just that important, and Google Maps, how flawed it can be, is currently the best for so many areas.
You'd pay 50-100 euro a month ? Why ? How did you arrive at this number ? Even before google maps, the standalone navigation devices didn't cost that much. To stretch this, if maps is 50-100 a month, how much is the total bill of all software you use ? Surely, the h/w, OS, browsers etc. should be orders of magnitude more than what they are today too.
Straight comparison to the other service I paid for for a while I gave up as it didn't seem worth it.
> Even before google maps, the standalone navigation devices didn't cost that much.
They costed a bunch and were a lot less useable in day to day life.
There was no device that would tell me the schedule of the next train departing from the station 10 min from where I am, when the store I am going will close, and how I'll get back when there's no train left for the day anymore. To your point there were service providing this info in part, in text format, on super small screen, and they'd cost around 5/10 bucks a month.
Now I get that with a mostly accurate map, end to end, automatically udpated as I start moving. And I also get other people's position, and don't need to text them every 5 min to know if they're lost or stuck.
The same thing goes for mountain maps, where I live the trail markers are relatively accurate and I can use it as a reference to check the official maps and on trail directions (and getting the other people's position is that much more valuable)
All in all it's a package that was never at hand before smarphones arised, and is currently only really working with Google Maps, even as we have OSM and Apple Maps at the distance.
> how much is the total bill of all software you use ? Surely, the h/w, OS, browsers etc. should be orders of magnitude more than what they are today too.
I actually think that keeping a worldwide map including cities street level details and natural paths up to date could cost a lot more that maintaining a browser or an OS.
> There was no device that would tell me a) the schedule of the next train departing from the station 10 min from where I am, b) when the store I am going will close, and c) how I'll get back when there's no train left for the day anymore. To your point there were service providing this info in part, in text format, on super small screen, and they'd cost around 5/10 bucks a month.
There is a free Android (+ iOS) app doing a) and c), it's Transit App, at least in 18 countries and 300 cities [0]. Gives you schedules and (multimodal) trip planning including car, bus, transit, train, subway, cycling, walking, rideshare, bikeshare, scooter rental. It has both scheduled and (near-)realtime information, detours, service interruptions etc. Shows you all the options you're interested in, rolled up by start time and total time, and includes cost(/estimated cost for rideshare).
In text format, on a super small screen, free.
> And I also get other people's position, and don't need to text them every 5 min to know if they're lost or stuck.
Well ok but very few people use that much. Why do you need it so much? Socializing? Hiking? Playing Pokemon?
Transit looks like a nice app. I assume it's based on the open data initiatives in places where it works (France etc.). There's similar iniatives in many places now (Japan has one as well) and I'm pretty much rooting for more of those.
You're still left patchworking all these apps, switching from one another at every point and assembling everything in notes by yourself. It's super nice it can be done at all, but I'd also pay for an app that has it all in one place if it wasn't free.
To note, the c) part is not just public transport but anything from private bus network/taxi/private bike networks/just walk. While Maps deals with that decently, I'd agree it would be way too much too ask from a free transit app.
Transit App works fantastic in urban US, Canada, Western Europe, Aus, NZ. (It's also great when visiting a new location for discovering which routes get you from A to B, speed/frequencey/cost, as opposed to commuters who already know which mode, route and stop they want, and primarily want real-time updates, ETAs. Obviously you can supplement it with the local transit app/website.)
There's not much patchworking/assembling once you bookmark your preferred locations, research what times they open/close, select favorite transit modes, bookmark specific lines, research which exact stops. Don't mentally lock into Google Maps as your primary thing.
I said the c) part is already done by Transit App and generally better than Google Maps (shows rideshare time-and-cost estimates, bikeshare/scooter pickup locations etc.) Where 'taxi' has generally been replaced by rideshare and by 'private bike networks' you mean bikeshare. Not sure if by 'private bus network' you mean coaches, or employee shuttles, but those are sometimes not open to the public. (Can you cite us an example location and name the private network companies?)
As mentioned, the amount of local detail depends on how much your municipality participates in open-date initatives; if it doesn't, ask them to. Transit App is adding new regions constantly.
I'm not the commenter you're replying to, but Google Maps was the sole reason I changed from a feature phone to a smart phone. For people who travel a lot, it's hard to compare life before and after digital maps. They are extremely useful.
50 eur a month is insanity. I bet that i and all the other 'poors' like me would start buying $5 permanent paper maps and asking for directions again. That's the kind of hemorrhagic price that only someone who never thinks about the cost of groceries or gas could pay. I'm not trying to tell you that you're a rich asshole or anything, but i do want to communicate that those numbers are vastly detached from the reality of most people. I'd pay 50 eur one time but monthly is a lot. Just pin me to the version i paid for, don't update road closures or any of that other bullshit and let me opt out of the parasitic subscription pricing.
I get your point of view, and I think it comes down to how much you rely on the service
As you mention gas, for instance my gas cost are ridiculously low (less than 10 bucks a month) as the car is only for specific cases, and 99% of the time I'm not driving.
Same way I care a lot more about visiting new places than going to theaters, and I'd never end up with 5$ of paper maps (to dig my own grave, I also buy paper maps of the wilder areas I go and they sure cost way more than that...)
On the other had, getting lost somewhere I have no familiarity with, being stuck at a station that has an incident and not having any backup plans, looking for someone in a town for 30 min because they're describing places in a way that is too vague are situations that are highly stressful and can cost a lot more than 10 or 20 bucks. Have you ever missed the last train and thought how much it will cost you to do something about it ?
That's where I put the price of an app I actually use a ton, it could be one of the main reasons I have a smartphone and not a combination of a dumb modem and a tablet.
What kind of comment is this? You act like the commenter has insulted you personally, just because they appreciate the service and would pay a lot of money for it. You even go as far as calling the other person an asshole, in a weaselly way. Each one of us have different means and put different value on things. People pay €50 to do a bungee jump.
That’s a stretch but I agree id pay 50 bucks a month for a total Google Subscription: gmail google maps docs and drive. All of it. No ads, personalized only to serve me. And bring back reader and I’ll pay 100.
The only case I can think of where google is the best is traffic updates. I use OSM primarily, osmand particularly for navigation, and I've tried going back to google a few times, every time I try it it's worse than the last time.
You may be willing to pay €50 a month for it, but most people can't afford that and wouldn't pay it.
Google Maps has its pitfalls, but it's by far the most up to date in the cities I've used it.
OSM lacks a lot of details that completely change how you'd move from point A to point B (I was missing overpass bridges and crossable paths in residential neighboorhoods), and point to point navigation is also not great. Then it completely lacks all the more commercial features Google has built on top of Maps (store details, position sharing, timeline etc.)
On sheer readability I also prefer Google Maps, but I've been using it so long that that might be just familiarity.
I wish I'd like OSM a lot more, but every time I ended up back to Maps for a reason or another.
> On sheer readability I also prefer Google Maps, but I've been using it so long that that might be just familiarity.
Interesting – personally I find Maps absolutely horrible for usage as an actual map for orienting myself (as opposed to just serving as a vague geographical background for displaying POIs or navigation information).
When you zoom out, it often doesn't really distinguish between forests and other open spaces (admittedly OSM's coverage in that regard varies regionally, though at least in Europe it seems quite comprehensive and definitively better than Google's).
Then, when you zoom in everything just turns into a featureless grey-on-grey with no distinction between built-up areas and everything else (only "parks" get shown, but e.g. in France that apparently even covers large scale "natural parks" covering hundreds or even thousands of square kilometres, so in that case everything, including any towns happening to lie in that area, just gets shown with a green background, which is equally useless), and buildings are only shown when you start zooming in really closely.
Yes, there is a lot to improve. I also have my gripes about what POI are kept from the detail view to the zoomed out view, or how I'll completely lose a location if I happen to misclick on some random POI that happens to fall under my finger on the edge of the screen.
On the rougher part of the maps, I often get back to satellite view and/or StreeView if available (even as they sometimes don't show the same info as they come from different points in time -_-;). It's a handy backup that I don't get on the open source map apps
> even as they sometimes don't show the same info as they come from different points in time
And for unfathomable reasons Google Earth (the desktop version) can show historic aerial imagery, but no historic street view pictures, whereas Google Maps works just the other way round – you can view historic street view photographs there, but no historic aerial/satellite imagery. (And the browser version of Google Earth apparently can't do either.)
For me in Europe, Google Maps coverage quality can be best described by this personal anecdote.
I used to live in Nürnberg, which is in top 10 German cities by population, and where are some major and well-recognized international companies are headquartered.
Nürnberg has a subway system (U-Bahn) since 1980s, and it's significant enough: a few dozen stations over three lines (one is fully driverless, btw).
Google didn't have any representation of U-Banh in Nürnberg till at least 2017.
I don't mean "wasn't supporting it in navigation and routing", I mean " stations weren't even marked and labelled on the general overview maps.
And it's not like they didn't have the into: there was a widely-used user layer which added at least station labels.
They just simply didn't care enough, and had other priorities.
In the meanwhile, the level of detail on OSM covered details as minor as every mailbox not only in Nürnberg, but in every small town around Frankonia (I used to participate in postcrossing and used this a lot from random places).
As a counter argument, here in Sweden I've yet to run into a place where Google maps has failed me when it comes to driving, public transport or address/POI search. OSM on the other hand is missing half the buildings once you get more then 15 km out from major city centers and even in major cities, things like house numbers and addresses is often wrong. The only scenario OSM is better than Google Maps is pedestrian and hiking routes.
France and Spain had a pretty good coverage, Japan cities are decent as well. I've only seen Munchen and Kolhn's most touristic areas but there were decent enough we didn't hit any critical issue. Commerce data and opening hours was abysmal on the other hand.
OSM was pretty good too in France but has different issues: they don't get the same access to up to date commerce data as exposed through local aggregators, and there's just not enough user data to have good heuristics on navigation times.
Depends on what you are looking for and what country you are in. When it comes to small 'unofficial' roads, bike paths and hiking trails then OSM is much better. When it comes to finding the location of rural buildings, addresses and companies I find Google Maps much much better and complete, at least here in Sweden.
Yes. On the more critical part when in the middle of nowhere I often use a traditional map as reference after getting my rough position through Google Maps. I wouldn't trust any of the mapping services more than the local entities to provide accurate trail info, especially as they're the own maintaining the paths.
Sadly I'm still not good with really basic compass and map positionning, and the GPS + average location info helped a ton in the past.
I don't know about "most" but many people can pay 50 Euros per month for a service they care about. The problem here is that very few people would pay a monthly fee for a maps service that was free for almost 20 years.
Monetization was a dismal failure. I don't know how well they're doing now, but Maps was a gigantic money-loser, forever. I'd be a little surprised if it didn't still lose money, but maybe less. I don't what those "pin ads" cost, but I'd bet it's way less than a search ad.
If you don't believe that, that's fine. "What about indirect revenue?" you ask? Google consciously does not want to estimate that, because such a document could be discovered in patent litigation. As it is, there are tons of patent lawsuits about Maps, and the damage claims always tried to get at Ads revenue, because Maps revenue was nil.
Caveat: I could be way out of date here. I've been retired a while now.
As for the UX: "enshittification" and big-company bureaucracy describe it pretty well.