I don't know if Google needs a new CEO, but from everything I've heard about internal Google, they need to change up how they measure success and who they reward. So many awesome products launched with tremendous hype, left alone to stagnate, and then unceremoniously shut down. At this point, I can't get excited when Google announces something, no matter how amazing it seems.
At this point, they're doing a very poor job of maintaining their core, legacy products. Search, their bread and butter, has been going downhill for years -- and is practically unusable today.
I keep seeing this sentiment around here, which confuses me because I think gsearch is better than ever. I find what I need instantly in 90% of cases, and when I don't, other engines don't really do better. Granted, I use ublock, but even with that turned off I'm not bothered by the ads. Can you provide examples of a query which gives unusable results, but Bing (or whatever) works?
I think the comkon case is if you're looking for anything niche/technical, it'll often just reinterpret your query as if it were a typo even if you quote it. That and searching for product reviews usually just returns SEO'ed trash like 10 10 lists or best vacuum cleaner 2020 that's just a mishmash of press releases as opposed to actul experience or useful information.
How many people and organizations online actually provide real filtered hands-on reviews of things these days? There are exceptions like Wirecutter and sites like Amazon provide some insight assuming you're willing to wade through all the fake reviews, But generally speaking, consumers aren't willing to pay the cost for in-depth reviews.
ADDED: There are individuals in some domain niches but you pretty much need to know who they are.
Google was sitting on a massive cash cow, so they appointed someone who would just keep it ticking over.
In the face of an existential threat, like ChatGPT, they need someone who can actually drive innovation. Not innovate themselves, nobody expects a CEO to do that, just create a culture which has a hope in hell of rising to fend off challenges to the empire.
They don't have this.
That have a CEO who is only capable in "good times" ... along with most of the company.
Faced with sufficient adversity, they will need a CEO who can succeed in the face of adversity.
Reddit has lots of them. Not sure why google can’t pick them up. YouTube also has specific reviews on anything you might want. Considering google owns YouTube, it’s strange that these results don’t bubble to the top.
YouTube has ostensibly strong policies around disclosing commercial activity, but Reddit does not. This "you can trust Reddit for reviews" narrative is nonsense. It costs basically nothing to subtly shill a product on Reddit, and more sophisticated operations will manipulate upvotes as well. On most niche subreddits, you only need a dozen upvotes to bubble to the top. Impossible to track abuse on that scale.
Not to mention the transparency nightmare of subreddit moderation.
Now, you're talking about essentially manual curation of trusted reviewers though. At that point, you're getting close to essentially resurrecting a Yahoo-style directory of good content. I don't really use Reddit but I have sites and people I go to for reviews of certain types of gear. But I don't know how scalable that is.
A big part of Googs secret sauce is ranking relevancy in part by reputation. They totally could index text-to-speech of youtube videos and rank channels by popularity in their content niches, then supply those when searches overlap their the content terms. Clearly they do _something_ like this, and probably could do more. Wirecutter and reddit should show up higher than random SEO fake-review sites, not because of manual ranking, but because NYT and reddit get higher traffic, and have higher reputation from other sites. They should able to derive signals that them so.
For Google, given the amount of knowledge it has on people, it's possible to know if someone is reviewing a product just for the sake of it, or is it a job for him.
> it'll often just reinterpret your query as if it were a typo even if you quote it
This is exactly it for me, there's no way to tell the computer "this is what I want you to do." Maybe it's because product designers don't want to tell the user no. If something does fail to work, we couch it in language like "oops, something went wrong, we're very sorry" with no further hint of what to do next.
Sometimes I just want the big database in the sky to tell me what it knows about a specific string of characters; this seems like it should be straightforward but I suppose it isn't.
In general, it's probably true that the amount (and sophistication) of spam and spammish content has tended to overwhelm whatever technology improvements in search there have been.
It's just that to the HN audience, the spam sure seems unsophisticated. I can imagine being able to filter out most blogspam just by looking at the CSS. like, every blog spam review and restaurant site is so obviously bad it hurts.
also, "bestmitersaw.com" or other similar single product review/spam sites filled with affiliate links. Jeez. Kill them all.
Also feel like if there was more then one way to interpret a query, it used to try and give an example of each on the first page. Like it used to be if I searched "ford", I'd get both links relating to Ford Motors and President Ford. Trying today, the first link to Prez Ford is on the third page.
I don't think you will ever get a 1:1 answer of a query that "works" in DDG/Bing but not Google, search is too curated. From my experience, when I am searching for what seems like a rare - but not completely unheard of errors happening in my code, on my sprinkler system, anything, it seems like the Google results are just crap compared to what the same query 5 years ago would have yielded.
I don't know what it is, if their algorithm is getting worse, or there is just too much shit on the internet to accurately index. From my experience, Google Search hasn't even just stagnated, it has gotten consistently and steadily worse for at least the last half decade.
Same thing happened to my Google Home, literally got dumber as the days went on, queries that it could answer yesterday, would just stop working the next day. I threw that thing in the garbage.
However once you look for very specific narrow results it attempts to outsmart you can be a real annoyance. When it decides to ignore some rare term and correcting your query while that rare term actually is the thing of highest significance.
So yeah, for 95% of the searches it is great, but on the 5% it feels like modern advances are in the way.
If 95% of the Google searches I made were great, I wouldn't have stopped using Google. My experience is more like 10% of the Google searches result in something close to what I'm looking for before the third or fourth page.
Can’t say much about other search engines, but Google has definitely let it slip really bad. For example, organic results being pushed below the fold, with organic-looking paid results taking top spots instead. Another big issue they’ve had for about a year now is in my native language search, the organic results are completely been taken over by Google-translate generated content that actually redirects to random aliexpress items. Their language models seem to prefer Google-translated sites over human generated language for some reason.
I agree with you that saying it's unusable is ridiculous but I HAVE noticed a marked decline in results. I still get what I need 98% of the time, but now that 2% has gone from no results to having to figure out how to force Google to search for what I want. I'll search for a phrase and frequently it'll flat out ignore a word I entered, so I'll put it in quotes, and it STILL ignores it. Then I'll result to an entire phrase in quotes and I just get trash results. I feel like advanced search features have been nerfed for some reason.
I find that if I'm looking for a factual topic I just get heavily SEO'ed, heavily ad laden blogspam. The algorithm must be weighting dwell time because if you are looking for recipe for how to cook eggs you'll get articles with 7 pages of ads interspersed with fluff filler copy before any actual recipe. And they probably have no incentive to cut this down because much of those ads might go through google's own ad networks.
The recent yandex leak had dwell time in it, I think. It makes me sad that it means that recipes (a very competitive field) are going to be coming after novels on the top sites.
From personal experience, Google search on a desktop w/ Adblock is great. And especially when you know what you're searching for, the experience is usually great.
On mobile or without adblock though or when trying to do discovery of something new, you really have to sift through pretty much the first 5-7 results as they're all sponsored, ads, or terrible SEO spam.
But it is a hard problem to solve, and no other search engine is better in that regard imo.
It can filter out native search results as well, as long as you tell it what to filter out. It's of course a bit limited since you have to know beforehand what you don't want to see, but it works great if you're looking to, for example, filter out specific websites.
It can, but I don't know of any default filters that would do such a thing, and that is besides the point. It certainly doesn't have enough time to classify a particular entry as SEO spam without doing RPC to an external service, IMO.
My suspicion is that most of the people that feel it is getting worse either a) hate ads or b) don't realize they do harder searches than what they used to do.
I'm fully aligned on hating ads. It is silly how the top results are all sponsored. Curated, I would like. Sponsored? Not so much.
My searches aren’t harder than they used to be. If anything they’re easier because I often know precisely what I’m looking for.
I stopped using google because it ignores my search terms and I find that infuriating. I’ll use a lower quality search engine simply for the lack of annoyance. Whenever I open a new chrome profile I’m reminded that I objectively hate google search.
> they're doing a very poor job of maintaining their core, legacy products.
My guess is that great PMs and engineers are not into incremental changes or maintaining a mature system. As a result, the sharpest minds for building product would have left the search org over the years, therefore the quality of search deteriorates over time. I'm not saying search org does not have great talent, but I guess their talent is about rearchitecting search infra for the 101 times for no groundbreaking gain.
A lot of the allstar PMs and Engs who helped actually GTM a lot of great Google products like Maps, Docs, Ads, etc left in the 2010s to start their own startups (eg. Jens Rasmussen - founded Google Maps and then left to found Where2), become execs at other companies (eg. Niklesh Arora @ PANW, Lars Rasmussen @ FB, Bret Taylor @ FB+Salesforce), become VCs (eg. Bret Taylor, Niklesh Arora), or (re)enter Academia (eg. John DeNero @ Cal).
There also was a change in how PMs began getting hired at Google in the 2010s when they brought a senior ex-McK TMT consultant to lead and revamp the Product team after Marissa Mayer left to become CEO of Yahoo.
This is part of it. It's also well understood within Google that the way to get promo and move up (and secure your job...lack of 'upward momentum' means you'll be first on the chopping block) is to launch. Getting 'stuck' doing maintenance and incremental improvements is a death knell for your career. Thus the pattern of big flashy product launch, followed by stagnation and eventual cancellation.
Instead, my guess is that they are optimizing the system for maximum ad revenue--and that the products only have to be good enough so users won't switch to Bing, etc.
> Search, their bread and butter, has been going downhill for years
As a user, yes. But based on the extra ads and garbage I suspect that profitability of search is at an all time high. So as long as search keeps “growing” then there’s an opinion that it’s a success. I think that’s dangerous for long term thinking and will lead to Google’s demise (see IBM, AT&T, Comcast, etc) that were profitable with horrible UX, until they weren’t profitable.
AT&T didn't get outcompeted. They got broken up by the government and before that were forbidden from commercializing discoveries some of which weren't really commercializable (like the origins of the universe).
I have been using a different search engine for about a year now and almost never go back to Google anymore, and when I do go back to Google it's almost always to use it as a dictionary (search: define something).
Your describing the outcomes of the old "perf" performance review system. Cultural issues like that are hard, but not impossible, to change. GRAD changes things, but as long as promotions are tied to shipping new projects and not excellence in maintaining those launches we will continue to see this behavior.
GRAD in theory represents a huge improvement on Perf. Rather than having to please a faceless committee by stopping the whole company for several weeks per quarter to write performance data for the committee, you now have to please your manager and it's their job to gather that data.
The layer of indirection between you and the calibration committee in Perf meant that maintenance work, since it does not generate nearly so many “artifacts” (design docs, engaged user statistics, lines of code written, code review requests) it was nearly impossible to quantify and reward. Your best chance was to ask people on neighboring teams to give you Kudos or Peer Bonuses for it, so that the committee would say “well it must have been important they got all of these gold stars!”
This is how you get a dozen competing chat apps, “improved the last chat app” cannot appear easily on Perf, “wrote a new chat app and we already have 50 users!” can.
However, though it affords an opportunity for things to be better, the internal rollout of GRAD has also been botched this past year. In the absence of strong leadership, managers are assembling the same data as before “we just must not call them perf packets”, and potentially reinstituting the same biases as before. (I was actually told by my manager that I could not get a rating above a certain ratings ceiling even though he wanted to recommend me higher because his director phrased a general rule that would apply to me, “if X then they should not get a rating above Y,” but hi I am also X on something of a technicality—and my manager was not sure he would have the political capital to argue my case.)
The botched rollout means that these culture improvements from not having to prioritize only what can be easily quantified, are delayed. Like, seriously delayed. Imagine that with really strong solid leadership you could have changed the culture in one year, but now that people are reconstructing the broken familiar process inside of the new process, the culture is not going to change for the next 5 to maybe even 10 years as second order influences slowly nudge it into a better scenario.
Google leardship was primarily reacting to threats, such as Amazon S3 and the iPhone, rather than proactively innovating. Despite the fact that AI is widely regarded as the future, Google has not made a significant effort to dominate in this field.
I think Google needs a visionary leader who can anticipate the future and make strategic investments. I do not think internal structure will change without visionary leader.
> Despite the fact that AI is widely regarded as the future, Google has not made a significant effort to dominate in this field.
I think of “Google and AI”, and I remember hearing more about their work on “AI ethics” than work on actual “AI” itself. Of course, ethics is important, but if you are more focused on the ethics of doing something than actually doing it, that’s scarcely a recipe for leading the market-on the contrary, it sounds rather risk-averse-and, maybe that’s unfair, I don’t know, but that’s just the impression I get from my own limited perspective.
Google uses AI/ML all over the place to incrementally improve their products and always has but what people mean by "AI" in this context is the big LLMs or generative AI.
> Google looks like the IBM to their Microsoft right now.
I really dislike how Microsoft is being considered to "be better at AI" recently. No, they just integrated another AI solution in Bing. They haven't released any big AI product themselves for a long time.
> has not made a significant effort to dominate in this field
It's almost like Google is not immune to the fate of being a massive company.
And just like prior large, market dominating firms (IBM / Microsoft / Intel), managed to miss a new trend (commodity PC hardware / web/mobile / contract fabbing) due to wildly profitable legacy revenue streams whose sheer size masked competitor acceleration.
> Google has not made a significant effort to dominate in this field.
What? Tensorflow, one of the 2 main frameworks for doing AI work is a google project. Google has domain specific processors to accelerate AI workloads - tensor processor. Google Brain. Id argue they have been making the biggest effort out of all the large tech conglomerates. They just haven't made an effort to turn their developments into consumer facing products, instead of milking the ad-cow even more.
Google has shown they can build amazing AI that can play chess and go, nice … then what? They built borg and gave k8s to the world and then what? OpenAI took the stage on ChatGPT and AWS dominates the cloud.
This goes back to way promotions are handled: clearly people are incentivized to build stuff but no one cares if that would have any path to profitability.
Does the world really need fucsia? ChromeOS?
A business needs both and needs to incentivize innovation along with long term commitment.
kubernetes came after the rise of aws, likely to keep google in the game
> Does the world really need fucsia? ChromeOS?
ChromeOS is an extremely popular OS, especially in education markets and there is definitely a need for such an OS. The majority of computers in the world are overkill for what the average end-user is trying to achieve and chromeOS fills that space very neetly.
Otherwise I agree with you that technology needs more than just technical vision to succeed, but my opinion is that AI isn't ready for the mainstage and maybe some people within google thought the same and chose to not rush anything. Then google made a massive blunder by going on the defense with their bard unveiling.
I don't think it's the problem with lack of innovation. Google has innovated in a lot of areas - Transformers (the foundation for ChatGPT), Stadia, Project Starline, and some the public may not have heard of yet.
The problem lies with translating that innovation to long term product vision, and reimagining how these innovations could be made into a product.
I suspect that they announce it to fast too, too soon - before it's proven. Stadia and LaMDA is an example of this. They actually showcased LaMDA last year or the year before that in I/O. The other is that they give up on innovations too soon because they can't think of ways to monetize them.
Pichai needs to go, and I've been calling it here for over three years.
It's been so obvious Pichai isn't the one to lead Google. Here's a mountain of evidence where you can see him failing to adapt to a risky situation and changing market:
Pichai has a history of not being bold, not responding to threats, not innovating, not fixing broken culture, not seeing the 20 year view of the market. He's lost some of the biggest advantages Google once held. He needs to be done. And he needs to be done three years ago.
Now when he's tasked with reversing the sinking ship? Totally fumbling.
I agree, but not with the conclusion. I don't think a Nadella is what they need. He's far too much of a stabilizer, which makes sense for Microsoft given their incredibly rocky years. This archetype is also what Pichai is trying to fill and why Google isn't working out.
When you step back for a moment, it's utterly bizarre Tim Cook was able to pivot Apple towards being this privacy-focused company who is now able to eat Google's ad lunch from right under them. Tim Cook, the very same guy people called just a boring distributions-focused leader, yet he has a more forward-thinking vision than Pichai. Pichai, the man who once led a team to create Chrome.
It's outright absurd how much one has stalled in comparison.
What I like about Nadella is that he is very responsive to threats, and I think he fundamentally gets the business. Much of MS' product line is improving across the board and ramping up vs AWS.
Cook really understands the customer perspective around privacy, and I view him as part of the A bomb that will destroy the glacier of Surveillance Capitalism. FB is limping, and Google is bleeding. He'll likely be ramming it at Mach5 with the subscription icebreaker which feeds right into the Apple ecosystem. Now, I do think he needs to drop %s and bring them down to reality as part of a long play to make it the best platform and defeat monopoly arguments.
He was a PM but there's relatively little PMing to do on a browser. If you look at the features Chrome launched with it was all engineering focused stuff: speed, security, non-functionals like that. The project itself was really driven by Page and some hires from Mozilla/Safari teams.
As if Google has half the staying power of IBM. You have to eeek along for 50 years to be IBM. Google has 1 cash cow and if it does it will be more Nokia than IBM
Id be totally willing to see Jeff Dean promoted to CEO. Not that I think he'd do a great job, but it would be interesting to watch. He's been building a case for this for quite some time.
> Eric Schmidt is busy trying to fix Pentagon software development,
That could actually play to his advantage if he really wants to be CEO again, it is my belief that in the current geo-security climate all the big US tech companies will become more and more entangled with the MIC (transforming it into a MIC+tech thing, in a way).
But I suspect that at this point Schmidt is trying to play the big political cards. Maybe he wants to become the new government-appointed tech Tsar in case that job will be created at some point? (I personally think that it will be created).
Nadella has done a better job than Ballmer but don't overrate him too much! He has many major failures on his watch. Microsoft has been very effectively buying a warm reputational glow with developers by splashing money around, but that's all relative to their previous state. It's still a company with a lot of problems:
- With the exception of VS Code, the Microsoft things developers like are purchased or just catching up with where competitors were 20 years ago e.g. GitHub, Minecraft, OpenAI, Windows Terminal, WSL.
- There's a top thread today on HN all about how much people hate Teams. Teams is an abomination yet forms the central pillar of their collaboration strategy; how can they have executed so badly that people seriously say they'd rather quit their job if it's introduced than deal with its terrible design?
- Windows has totally rotted away on Nadella's watch. The execution failure here is total. Does Nadella even use Windows himself? It's still critical infrastructure for much of the world but the entire Windows platform has no vision, no ideas and has become (just like Teams) mind-blowingly buggy. Absolutely nothing they introduced in the last 15 years or so actually works properly. A good measure of the totality of Nadella's failure here is that Microsoft's new apps are all Electron based and Edge is a fork of Chrome i.e. they institutionally prefer to use Google's code to their own. Also they apparently hand out MacBooks to new employees!
- Visual Studio was once dominant in its space but is now apparently considered a writeoff, so VS Code is a totally different app written from scratch.
- .NET has experienced major product and branding churn that has left anyone not fully committed to the ecosystem confused about what it is and can do e.g. .NET Core vs Framework. .NET isn't a failure but it doesn't seem to be growing or exciting devs either, and in terms of core R&D they've been blown away by the work Oracle is doing with Graal.
- Azure is very successful but is showing worrying signs of the same bugs-and-tech-debt problems that plagues the rest of their products. When I had to use Azure, I was constantly amazed at how even basic things were flaky or did not work. Like even just holding open TCP connections between machines wouldn't work properly. This is despite the fact that Microsoft was able to built out Azure seemingly only by raiding the rest of the firm for talent, hence the Windows brain drain. Also cloud is becoming strongly associated with cost disease, which opens up a possible weakness in future (in fairness, a problem shared with AWS and GCP).
- HoloLens is cool but does anyone use it?
Nadella fixed the monolithic our-way-or-the-highway approach, which allowed them to import tech from elsewhere to fix their weaknesses. But compare that to Apple, which consistently doubles down on its own tech stack and has become even more successful with the monolithic integrated approach and it's hard not to see this as something Nadella was forced into rather than some great act of leadership.
I don't think this is the right way to look at it at all.
> With the exception of VS Code, the Microsoft things developers like are purchased or just catching up with where competitors were 20 years ago e.g. GitHub, Minecraft, OpenAI, Windows Terminal, WSL.
Yes, but now they're leaders in every one of these categories. And instead of holding back acquisitions, Microsoft has made each one demonstrably better.
> how much people hate Teams
Teams is a beachhead, and they're absolutely going to take it and win.
> Windows has totally rotted away on Nadella's watch. The execution failure here is total. Does Nadella even use Windows himself?
There's really not much need for Windows anymore. Putting a veneer on Linux would do just as well. That said, they're continuing to win gamers and over half of businesses. The future is thin clients, though, and that's where the battle is shaping up.
> Microsoft's new apps are all Electron based and Edge is a fork of Chrome i.e. they institutionally prefer to use Google's code to their own
The entire world chose Blink/Chrome/Chromium. Keeping their old browser around would have been a major mistake. And now they're set to take the the lead. Google will be too busy defending their own moat to keep up Chrome leadership, and Edge will push search margin to zero.
> Visual Studio was once dominant in its space but is now apparently considered a writeoff
The way people write software is changing. Cloud based thin clients are what businesses want. The entire development setup virtualized without DevOps. Fully secured. And beyond that, AI-assisted coding. Of course Visual Studio is a dinosaur.
> .NET has experienced major product and branding churn
Microsoft is busy winning Linux developers and the rest of the world.
> Azure is very successful but is showing worrying signs of the same bugs-and-tech-debt problems that plagues the rest of their products.
It's a top cloud contender and will gain even more market share.
> HoloLens is cool but does anyone use it?
A small, synergistic bet on spatial computing and XR that keeps them in the game should the field pop off.
