9 ms·

Ask HN: How did Google botch messaging/video/hangouts so badly?

166 points by davidw 5y ago

Several years ago, from within Gmail, I could: * Make a phone call * Send an SMS * Send a Google chat * Start a video call with my parents Now, all of this has been split up, and to start a video call, I have to start a video and send an email inviting someone. SMS has been split up into a separate web app that won't work without my phone being present. How did it come to pass that they took an easy, integrated system and mangled it so badly? I mean, it wasn't perfect, but it mostly just worked, and was easy to use.

pre 5y ago
Presumably they make more money with the user-hostile version and the user isn't paying so...
matv 5y ago
https://arstechnica.com/gadgets/2021/08/a-decade-and-a-half-of-instability-the-history-of-google-messaging-apps/ https://arstechnica.com/gadgets/2021/08/a-decade-and-a-half-...
- clircle 5y ago
  Incredible
noah670 5y ago
The real question is what newer products have managed to stick around at Google?
- imbnwa 5y ago
  It all went downhill once they shuttered Google Reader
- bitcuration 5y ago
  Gmail has. Knock on the wood.
  - vaseem 5y ago
    Actually it hasn't. Ever get annoyed whether your email is in primary, updates, promotions or social. Previously I had rules, almost all marketing went to labels or trash. Is it just me or have the rules become inconsistent?
    - mrweasel 5y ago
      I think I might be in the minority.I never use Gmail (GSuite, at work) via the browser, but I still use Gmail, it's just via IMAP or "native" Gmail integration. I was never blown away by the user interface, but many rely on the gmail interface to manage their emails effectively.
      - JohnFen 5y ago
        I don't use GMail for my personal use, but my work used GMail, and that's what I did with that. Google's actual interface is not one that I enjoy. Later, we switched to O365 and so use Microsoft for email. I still use IMAP for that, because the Outlook interface is not one that I enjoy.
      - dhosek 5y ago
        Me too. I have one account that I use for my fiction/poetry submissions that I'll log into via the web interface when I get a response because Apple Mail doesn't have a way to be able to apply multiple tags to a message and I tag each response with the journal, the submission and whether it's a rejection or acceptance (plus a star if there's specific feedback). Other than that, all my mail access is through Apple Mail either on my Mac or on my phone.
    - jeffbee 5y ago
      You can disable the categories if you care. See it is another symptom of the bottom-up product strategy. Categorization was a key feature of Inbox (by Gmail (by Apps (by Google (by Alphabet)))). After they killed Inbox they wrecked the categorization bundling, too.
netsec_burn 5y ago
I believe the reason is that there are incentives internally to create new products rather than maintaining existing ones.
princevegeta89 5y ago
The experience to setup and use was so confusing
- davidw 5y ago
  It wasn't, really. You had a Gmail account, and you could call/chat/sms/video call your contacts from within it.
  - tacostakohashi 5y ago
    I think part of the motivation is to provide standalone apps for people that don't have gmail accounts because they're using hotmail/yahoo/some overseas equivalent. Like you say, the old experience was great if you use gmail. Now the new split out messages thing is annoying for gmail users, but not tied to a google account. I think their strategy is to break things up into individual lowest common denominator services (like whatsapp), instead of having something that is pretty good at several things, but not the dominant product for any of them.
    - jakamau 5y ago
      But I can't think of a single messaging application google has created that doesn't require a google account to use.
mdorazio 5y ago
Having talked about similar questions with friends who work/worked at Google, you need to first ask "How do people get promoted at Google?" The answer to that (by launching new things that get abandoned soon after rather than improving/fixing existing things) answers your question and many others like it.
- amelius 5y ago
  And there's never another employee saying: wait a second, that's not a good idea ...?
  - davidw 5y ago
    As per my comment below, it's not just stagnation - I was ok with where they were 5 years ago - it's that they've made an active effort to degrade the experience. Instead of clicking once in Gmail to start a video call, there is now some janky system where it starts a 'meet' and sends my mom an email asking her to join it. An email! When both of us are sitting there with Gmail open.
  - infogulch 5y ago
    Which seems to be a reasonable human decision -- do you want to be the person that blocked your colleague from getting a promotion?
  - MisterBastahrd 5y ago
    That would require you to pretend that Google listens to devs more than the MBAs they work for. Nothing says "market expertise" like having a sheet of paper that proves you once completed a class project to write a business plan.
  - FinanceAnon 5y ago
    If many managers are onboard with an idea, one engineer saying "no" doesn't have much weight. Also, if you keep saying "no" to things, you might be perceived as "negative" and it might come up in performance reviews. And it's often difficult to predict if an idea will work or not - a lot of it comes down to execution and future politics. So, as an engineer, it's probably better to jump on the bandwagon with everyone else, and try to get a promotion from it. It's stupid, but it's how it works. You will not progress within a company by saying "no", even if you are right and it stops a bad project. (I've never worked for google, but did work in a large software corporation)
Gualdrapo 5y ago
I can't understand why they ditched XMPP for corporative Google accounts. Our uni uses Google services but for their stupid chat I have to keep a Hangouts tab opened because most XMPP clients won't work with it.
- mrweasel 5y ago
  While I don't know, I have a guess: Google loves building infrastructure, and they're pretty good at it. XMPP wasn't build in-house, and writing a replacement protocol was something Google could do. What Google absolutely suck at is UI and native applications (at least on the desktop). It's also why their products keep failing. Seriously how hard would it be to build native clients or open the protocol and have it integrated into libpurple. I find it extremely hard to comprehend why Google don't just hire five of the worlds best macOS developer, five Windows developers and five for Linux/BSD and just let them build fantastic clients that integrates into each platform and provide the very best chat experience.
- notatoad 5y ago
  IIRC they were pretty up-front about why they dropped XMPP - most of the clients connecting over XMPP were spammers, and there were few enough legitimate users connecting over third-party clients that it made more sense to drop XMPP support than build spam-blocking solutions.
- vadfa 5y ago
  Everybody I know moved on from XMPP years ago. Google and Facebook dropping their support was indeed big part of it. But all in all centralised protocols have proven to be easier, more stable and simply better. Want to roll out a new feature? No need to sit down with a committee for years or wait for all 3rd party clients to keep up: you just implement the feature yourself. Wait... what was the question?
  - Andrew_nenakhov 5y ago
    > But all in all centralised protocols have proven to be easier, more stable and simply better. Show me a centralised protocol that is better at interoperability between various services?
bitcuration 5y ago
The truth is, other than search engine advertising, google doesn't know how to make money.
- jeffbee 5y ago
  Close, but not quite. The only thing Google knows how to do is make dirt-cheap large-scale computers and hire people. Then they let the people do whatever they want with the giant cloud, because the marginal cost is nothing. So, the reason that gchat etc keep appearing and disappearing is because there never was a plan for it in the first place. One or a few engineers just threw it together at some point and there it was. Nobody truly owned it, it wasn't strategic. The churn is just a superficial symptom of the bottom-up product management style. There are tons of Google products and features that were thrown together by individuals or pairs of people over a weekend and then subsequently got launched to the public. That this is possible is pretty neat, but that the products might not be durable is a downside.
  - iamstupidsimple 5y ago
    > There are tons of Google products and features that were thrown together by individuals or pairs of people over a weekend and then subsequently got launched to the public. Do you have concrete examples?
    - jeffbee 5y ago
      I can think of two right of the top of my head, that were monograph side projects that got launched: Google Keep, and the comment system of Google Docs, which you can think of as maybe another chat system or social network.
    - ulfw 5y ago
      Remember Orkut?
  - MichaelZuo 5y ago
    ‘ because the marginal cost is nothing.’ Who pays the utility bills?
indymike 5y ago
> How did it come to pass that they took an easy, integrated system and mangled it so badly? I think each of these things were part of gmail at different times. Google talk (their original chat that competed with Yahoo messenger) was and sorta still is there even though the talk product is something else now. Google voice was required for SMS. It sorta worked from gmail, for a while. Then they tried to merge all the things into hangouts and it was kind of really nice for a few months... Then they tried Duo, Allo, Hangouts, and Google Chat (which is a slack work-a-likiesh thing) and I quite keeping track :-)
- davidw 5y ago
  For a while, it worked really well. I lived in Italy and had a Google Voice number from Oregon, and I could just make US calls for free from within Gmail and it didn't just work ok, it was pretty good. I could also place video calls to people who were contacts right there in Gmail with one click, and it worked well too. It's not that their system has stagnated - they've actively made it much worse and more difficult to use.
- jeffbee 5y ago
  Google Voice was not required for SMS from gmail. You used to just be able to "chat" with any phone number from within the gmail web frontend. It cost a dime, but your dime was refunded if the other party responded, as an anti-spam measure. It was also formerly possible to make a voice call to any phone from the gmail web frontend, no Google Voice account required. Now you need Google Voice to do this (not Google Fi!).
mdasen 5y ago
I'd guess that it's a combination of things, but I think what sticks out to me is that a) they probably hit a plateau and felt like they needed to do something to break out of that (and didn't really have a good idea so they started "mixing things up"); and b) messaging seems like a product that everyone wants to solve, but doesn't necessarily offer amazing monetary returns. I loved GChat. It was simple. I could use Pidgin or Adium or any number of third-party clients. It did what I needed. However, I'm guessing at some point, they were working on and maintaining a product that didn't really have the metrics that they were looking for. Growth probably slowed a lot once so many people were using it (so the honeymoon phase of "we'll figure out the money and strategy later" was over). Likewise, one of the great things about GChat was that it felt so un-monetized. There were no ads or anything. Compared to the alternatives of the day like AIM, Yahoo, MSN, etc. it just felt like this easy, clean, simple messenger that worked without distractions. When you're a company worth so many billions (now nearly two trillion), why are you putting engineering time into something that seems to have little growth and little money? Shouldn't you re-task those engineers to projects that might be the next big thing? I know that in a certain light companies can hire more people, but the hiring pool isn't infinite and you can only grow your staff so quickly without things becoming chaotic (you want enough veteran staff members around to mentor new people who don't know what is going on with the giant systems that are created in such a large company). To me, messaging feels like a product that everyone wants to solve because it's cool, but people haven't really figured out hoe to monetize it well. I think Facebook wants messaging to reinforce its ecosystem and fend off rivals more than anything. Apple really likes iMessage because consumers seem to be really into iMessage and it seems to create a positive feedback loop to get people to buy iPhones (I'm an iPhone user and I don't get what's so great about iMessage, but people are really passionate about it). What does Google get out of messaging? What does Signal get out of it? Signal and Telegram have both been looking for business models and they've looked into cryptocurrency, but I'd argue that neither has really found a business model. The messaging apps that seem to have found business models are the ones that aren't general chat/text replacements, but community chat systems like Discord and Slack. Microsoft's efforts with Teams and the new Google Hangouts Chat/Meet enterprise Slack clone show that Microsoft and Google see a Slack competitor as where the money is in messaging. It's easy to get a company to give you $5-25/mo per user when they're spending $5,000-50,000 per month on that user already (not just salary, but benefits, office space, equipment, etc). I think the real problem is that there's little money to be made in the old Gmail/GChat messaging. So, in comes some project manager that wants to make their bones solving a potentially large market in messaging and they don't have any wonderful ideas, but they're hoping that if they move enough things around and rebrand enough things they can cherry-pick some metrics and show how genius they are and why they deserve a big promotion. You don't get recognition and promotions for keeping a ship steady in calm seas. Combine that with a product that doesn't seem to meet expectations for return on engineering investment and why should Google keep investing in this? If we look at the companies that have succeeded, they're not general messaging apps and they're usually aimed at taking advantage of an enterprise play - with a generous enough free tier that home people can play with it. I think Google didn't want to continue offering a general purpose messenger that didn't have a path to profitable growth. They also didn't want to abandon messaging. So they kept shaking things up trying to find product market fit - profitable fit, not just something that free users enjoyed without something in it for Google.
criddell 5y ago
I've always wondered why they didn't just clone Apple Messages? Instead, they've done this seemingly random walk though the space and accomplished very little. Is it because Apple supports Messages through hardware sales and Google requires an advertising model? And things like non-encrypted RCS seems downright negligent. I don't know how anybody at Google working on this stuff can be proud of what they do. From the outside, it's a mess and I'd be surprised if internally things look much better.
boulos 5y ago
Disclosure: I used to work for Google. The chat and video call are in the same place again (both on the web and on mobile). The little TV symbol in Chat, invites someone to a video call. I personally don't love the WeChat-style "make the GMail app into your all in one comms app", but I get that some people do. The last part is that SMS and phone calls were in a separate universe and mostly focused on Android. Roughly at the scale of any mega company, it isn't one company: it's several that have the same funding and ease of transfer (people, resources, etc.) but not necessarily coordination. To wit, there are multiple "CEOs" (YouTube, Cloud, etc.) so I really do encourage folks to think of "Google" as like a dozen companies (Search/Ads, YouTube, Cloud, Geo, ...).
mikebonnell 5y ago
Are you sure you don't have Meet and Chat turned off within Gmail? If I go to Setting -> Chat and Meet within Gmail, I can have chats within Gmail. Meet still does send an invite/give me a link to copy, but I can paste into the chat and away we go in a new tab. (Former Googler)
obarthelemy 5y ago
I think it's - corporate ADHD. Managers whose project succeed are rewarded away, and the project flounders, until another bright young thing starts again... from scratch because ego - Google is all about algorithms not users, and messaging is mostly about listening to users and iterating. Google has thrown away... 5 installed bases of users ? - OEM politics: Google must constantly negotiate what is theirs and what is the OEMs'. OEMs want as much as they can, and that means messaging. It took iMessage becoming such a forte on Apple's side for OEMs to back off on Messaging and allow a single Android platform. - ditto carriers. - Apple is a spoilsport. Whatever Google do, on iOS it'll never be default, nor even as integrated as, iMessage. Knowing you won't really have access to the juiciest 50% of the US market is a bummer, even if you still can reach 80% globally. Who cares anyway ? Android allows one to use whichever app as default, just pick one. I actually removed Messages because it's idiotic and won't display a full text in the Notification, hence cuts off credit card confirmation codes, hence prevents me from buying anything from my phone. Idiots.
samstave 5y ago
The only really good service I like from google is MAPS
pedalpete 5y ago
I think this was the missed opportunity for Google+. If I recall correctly, when Google launched Google+, everyone we knew and wanted to communicate with was there. They had all the communication infrastructure, and even built it into the app with messaging and hangouts. But it was just mixed in with all the stuff you could do very well on facebook, and that everyone was doing on facebook. They were on to a good trend with circles, and that was a great differentiator from facebook, but they couldn't really tell us why we should use Google+. If they had focused down onto a niche in communication, I think they may have been able to crack the code. They would have been able to say "google+ is where everyone you know is, your friends, collegues, and family, it's the place where you can communicate with these groups independently." Unfortunately, they tried to be all things to everyone, and ended up being nothing to nobody.
foobaw 5y ago
reminds me of Google Wave
hbcondo714 5y ago
It's just not messaging / video / hangouts, Google "botched" their Pay app too: https://news.ycombinator.com/item?id=28281259 https://news.ycombinator.com/item?id=28281259
- chakkepolja 5y ago
  I wouldn't say botched for a single misfeature. It's pretty popular where I live, and I heard it's super convenient to do everything from phone recharge and electricity bill to train bookings. Whereas nobody talks about Google's messaging apps.
bythckr 5y ago
I really miss G Talk, it was such a simple & light app and it just worked. But I think the reason why Google didn't keep it is because there is no way to insert ads into it. This is the main part about Google that we have to understand - Google is an ads distribution company. Like a billboard company or the company that plasters ads on bus stands & public benches. Any product they have will have a place to serve ads. Serving ads in Gmail was a bad idea, but they have a nice user base. So, now they are offering it as a service that you pay for. I hope they come back to that. Your google id needs to be used as an id instead of relying on mobile phone numbers. I hope they bring back G Talk. The issue with meet is the hassle of sharing the code with everyone. They should just have the option of a friends list.

UncleMeat 5y ago

This is a meme. Yes, you need to change something about the state of the world but that change absolutely does not need to be launching a product. I work in the Core organization, which builds tooling for the rest of the company. My org within that has among the highest promotion rates (even to high levels) in the company and the primary thing we focus on is the sort of grungy "keep everything functioning" work that everybody says is anathema to promotions. It is no secret that Google's product strategy is a mess. But I'm not convinced this is caused by the promo process. For communications strategy I think it is instead the following: 1. Messy code stacks make surviving tool deprecation a problem. 2. Google seems to be bad at staffing for proactive protection against maintenance work. If a product is working, staffing dwindles to match the maintenance load. But if suddenly a huge mandate comes along like a stack switch, it is not easy to get the heads you need to handle that spike in maintenance. Instead, the product gets triaged or dies. 3. It is really hard to push past 1B users of anything, and growing populations outside of the US and Europe are not desktop focuses, encouraging a "mobile first" solution that leaves desktop integration in the dust.

contravert 5y ago
How many PMs does Core have compared to other PAs? The problem isn’t really with SWEs but with the entire incentive structure of PMs and the product management chain. They in turn are able to get more SWE headcount for these unfruitful ventures and end up failing upwards.
- UncleMeat 5y ago
  Core seems to have fewer PMs but we have similar problems with triaged products due to lack of heads. The PMs we do have also struggle with getting heads just as much as the SWEs do. At the very least, I think it is clear that Google’s product failings are not just a result of the promo system.
tdeck 5y ago
I'd suggest a 4th reason: Google has so much money and so many high-traffic properties that it can keep a poorly designed or unpolished product going for years, more than long enough for the original people who designed and built it to move on/up inside the company. Google will just push the product on the front page of some existing property to bump traffic, or get creators/developers to use it in hopes of boosting their existing search rankings, or give away the product for free. Every company would try to do these things but Google has a big advantage. This allows Google to be lazy about product-market fit and avoid asking the hard questions like "is this thing really better than the alternative" and "what's the ROI for people who adopt this". I worked on what I'd consider a failed product, and there was no explicit acknowledgement of that fact ever, we just waited 3 years and quietly turned it off. A more drawn-out version of the same thing happened with Google Plus. There was never anything like a "product postmortem".

vadfa 5y ago
Given that as of today xmpp practically doesn't interoperate with anything, well...
- Andrew_nenakhov 5y ago
  What do you mean, 'xmpp practically doesn't interoperate with anything'?! Dozens of thousands of xmpp servers interoperate and deliver messages to each other. Or do you expect it to interoperate with Telegram? What next would you expect, UDP <--> ICMP interoperability?
  - zaik 5y ago
    There are in fact transports for Telegram on the XMPP network. https://jabber.hot-chilli.net/server-specifications/#Transports_Gateways_and_Features https://jabber.hot-chilli.net/server-specifications/#Transpo... https://www.jabbim.com/home.html https://www.jabbim.com/home.html
    - Andrew_nenakhov 5y ago
      I don't consider transports as a kind of interoperability. From Telegram's point of view it's just 2 telegram accounts talking to each other. Wherever else this data is sent is not that important. Moreover, I came to a conclusion that reliance on transports reinforces the network effect of external services and weakens XMPP's network effect. Do not use them.
- MattJ100 5y ago
  There are at least hundreds of implementations of XMPP that are interoperable (at 5-6 actively developed server implementations, many more libraries and many client applications). There are at least hundreds of thousands deployed XMPP services. In addition XMPP can be linked with other protocols/networks via bridges/gateways. Some examples: - Spectrum: XMPP gateway project based on libpurple (support for many protocols): https://spectrum.im/ https://spectrum.im/ - WhatsApp bridge: https://git.eta.st/eta/whatsxmpp https://git.eta.st/eta/whatsxmpp - Telegram bridge: https://github.com/codingteam/emulsion https://github.com/codingteam/emulsion - Signal bridge: https://gitlab.com/nicocool84/slidge/ https://gitlab.com/nicocool84/slidge/ (replacement of https://gitlab.com/nicocool84/spectrum2_signald/ https://gitlab.com/nicocool84/spectrum2_signald/ ) Given such a diverse ecosystem of open-source and proprietary implementations, I'd say XMPP is one of the best examples of widespread protocol interoperability that there is.
  - fragmede 5y ago
    Good for them, but I can't talk to Gmail or Facebook users on any of them, which is the point. I'm not being flippant here, among the other things that Googlebook does is acquire users. Where number of users isn't a feature of the protocol, it's irellevant, but where that's a feature of why I would want to use a given protocol, it's the single most important thing out there. And XMPP fails miserably at that. Which is why open protocols don't work in practice. I love the theory that they would, and morn the days of Adium on OS X. But unless Facebook wants to build the bridge of XMPP to Facebook Messenger, the time I would spend working on getting XMPP setup is better spent elsewhere.