This always seemed like one of Google's worst decisions to me, and given that I remember the time they tried to spend $6 billion to buy Groupon, that's saying something. It was a plain unforced error.
Google Reader was a product that was crazy cheap to maintain (no way it had more than a dozen engineers working on it), and it was used primarily by the extremely online, especially journalists, bloggers, and other sorts of influencers. If they had viewed Reader as a marketing expense, keeping it online would've been a no-brainer.
But instead, they instead viewed it as a consumer tool that didn't have a path to profitability, and they were 100% right, but then over the next few years, it became clear how many later Google efforts would've benefited from Reader existing. Google+ would've meshed well with it. That little Google "stuff you maybe want to see" panel on Android phones that would frequently make little notices like "hey, we think you like this site, there's a new article," and that was probably really involved to build, and also it absolutely sucked compared to just having Reader.
Ultimately I have to assume that it's a business org problem. No engineer was going to get a promotion keeping Reader alive. No manager was going to be able to grow Reader's audience 10x. No director would want to give up a half dozen of their engineers to keep Reader running just because it would significantly hurt Google to turn it off. And no executive cared about a product that small. The org structure didn't lead to someone incentivized to want to keep Reader going, despite its popularity inside and outside the company.
> Google Reader was a product that was crazy cheap to maintain (no way it had more than a dozen engineers working on it),
It’s hard to show “scope” and “impact” by maintaining a product with declining usage. No one who cared about their career would want to work on something that was in maintenance mode.
I never understood why Google doesn't hire B teams to do the maintenance work, or outsource it. It's normal that if you hire only top players that no one wants to be stuck doing tickets.
Whoever's maintaining and updating Blogger should be involved in picking that team. People keep declaring it dead, and they keep putting out major improvements and modernizations at least every few years.
It’s not about hiring top players. It’s about the incentive structure of getting ahead. From all of the leveling guidelines I’ve seen at major tech companies, “coding quality” only comes into play to differentiate an entry level dev to a mid level dev.
I can’t think of any good measurable incentives that Google could use,
Rational beings typically act according to the incentives they are provided, so if coders don't want to work on Reader for this reason, and management don't want to keep it around, it does point to the incentive structure being part of the problem. But it therefore follows that many of Google's bad decisions could be partly traced back to how the company is organised - probably the interlinked combination of the incentive structure, and success metrics.
What's interesting is that even the most biased insider must see that Google really sucks in some domains - and recognising (and/or generating) user value seems to be one of them. Several times, they've either spent big on products which just missed the mark from day one (Plus/Currents, Stadia, multiple text/chat apps), or killed products which already had major engagement (Reader, Inbox, Orkut).
What's interesting is that if they made user engagement the only major metric that mattered and incentivised accordingly, it would likely drive interesting behaviour - a real focus on user needs, iteration to optimise engagement in the longer term, improvements in customer service, etc.
Because if the product is not in red (pretty sure Google Reader was in green), closing it burn user trust. I know about opportunity costs, but I'm not sure Google is in a position where it bleeds money so bad that need to rationalize things.
And having one more product helps you synergize strategies across the board.
99.9% of Google users are completely unaware of their reputation around closing down products. Closing Google reader a decade ago is a cause that nerds like us on HN care about, but virtually no one else. And the same nerds like us also care enough about our online presence to not use Google products for a hundred other reasons anyway
Does gmail make a profit? Does search make a profit?
Neither of them are paid services, but they allow a huge advertising business to be built on top. Reader could have been the same
Ok, that's great, because those are the people most likely to break a good thing. Maybe we can instead hire at least a few people (and given that Google uses shared database storage, build engineering, deployment, etc. I can't imagine we need many such people) who care more about things like a predictable schedule working on a low-stress product so they can have good work/life balance, to essentially just keep working software working correctly?
And then they get laid off and when it comes time for the behavioral interviews at their next company, they can’t discuss anything they did significant in STAR format.
And their skills are too rusty to pass the DS&A section of the interview and they didn’t do anything significant to show off their system design skills.
I work at $BigTech in the cloud consulting department (cloud app dev/“application modernization”) and the equivalent of “doing tickets” is being put on a project where you aren’t actually doing “consulting”, you’re doing “staff augmentation”. I avoid those like the plague.
Even though I’m not chasing after a promotion, I want to have something to show for the time I spent here when it comes to my next job either inside the company or outside.
That's the same problem, though, isn't it? Like, you are clearly not one of those people (based on your response there). And that's great: Google probably needs as lot of people like you to care to come up with new features and products. But the vast majority of jobs in this world--including ones that pay quite well, such as "doctor"--are not activities that are going to involve "scope and impact": they involve doing maintenance work using the established protocol--preferably with careful, predictable performance--on an existing system that someone designed or scoped potentially a long long time ago... one would expect most of software engineering would be the same way.
I guess the point I’m trying to make is that a lot of people make “enough” so that they feel like the extra stress of worrying about getting ahead isn’t important as long as they keep up with inflation. I’m at that point.
But, I’m not at the point where I can quit working. I’ve been around long enough to know that you always need to be interview ready when it’s time to look for your next job.
Being a “maintenance developer” is a sure fire way to get caught up in the “expert beginner” trap - “1 year of experience 10 times” and make yourself unemployable .
Yes but as with roads and other infrastructure, building is easy and shiny, but what makes them useful in the long term is maintenance. A very boring and seemingly unprofitable task. But the essence of being.
That is actually a very important observation about Google: products that Google engineers wouldn't elbow each other in the face to work on will eventually shrivel up and die.
Which describes for me everything that is wrong with Google these days. How can someone build a good product, if they are not even able to maintain a good product.
Google Reader was a product that was crazy cheap to maintain (no way it had more than a dozen engineers working on it), and it was used primarily by the extremely online, especially journalists, bloggers, and other sorts of influencers. If they had viewed Reader as a marketing expense, keeping it online would've been a no-brainer.
But instead, they instead viewed it as a consumer tool that didn't have a path to profitability, and they were 100% right, but then over the next few years, it became clear how many later Google efforts would've benefited from Reader existing. Google+ would've meshed well with it. That little Google "stuff you maybe want to see" panel on Android phones that would frequently make little notices like "hey, we think you like this site, there's a new article," and that was probably really involved to build, and also it absolutely sucked compared to just having Reader.
Ultimately I have to assume that it's a business org problem. No engineer was going to get a promotion keeping Reader alive. No manager was going to be able to grow Reader's audience 10x. No director would want to give up a half dozen of their engineers to keep Reader running just because it would significantly hurt Google to turn it off. And no executive cared about a product that small. The org structure didn't lead to someone incentivized to want to keep Reader going, despite its popularity inside and outside the company.