Hacker News new | past | comments | ask | show | jobs | submit login
Google Site Search will be completely shut down by April 1, 2018 (enterprise.google.com)
338 points by happy-go-lucky on Feb 22, 2017 | hide | past | favorite | 186 comments

I wonder if it would be worth the money for Google to keep services like this one running, just so that people (especially enterprise customers) don't lose faith in Google's competency with regard to long-term product support.

Although if you ask around here, you'd probably not find much faith for them to save.

I honestly don't think people outside the HN echo chamber care that much.

Problem is, people on HN are making decisions and building tools and startups. If everyone loses faith in Google, we'll stop using their platforms and tools which translates to loss of money.

The only example I can give is at my day job, I built a backup system that uses Azure blob storage because I actually trust Microsoft more than Google.

I wonder how many other people have a "anyone but Google" approach? It's not just cancelling services, it's the churn in APIs; they seem to be less stable than products from almost any other company.

Google clearly just doesn't have the enterprise-support mindset in their DNA in the way that companies like Microsoft do.

Google's core business is search and ads. Everything else is either an afterthought or a dalliance, and it shows.

The HN crowd is absolutely tiny, compared to the entire development ecosystem. And a lot of us aren't even shot-callers.

Are you suggesting there's zero people outside HN who avoid Google products due to legitimate concerns of long-term availability?

You should address that question to d0vs, as he is the one who made that point.

But now that I have a chance to think about it, I can see his point. Most companies don't really take that particular risk into account and just kick the problem down the road. If you tried raising it as an issue, you'd get responses ranging from 'sure, figure out a way to value the risk and we'll add it into the model', to 'who cares really?'.

This. Most corps with Deep pockets already have a big team on procurement, and I have no say in what tools I am allowed to use to get my job done. All those stuff is already decided. I may go and say, "Hey, have you seen this, it's better than X", and that will not yield anything at all.

I already avoid new Google products because of this, as I am sure many of you do.

While I've embraced Golang somewhat, it took me a few extra months of research and thought before doing that.

I can't wait for Rust to gain more traction, and be usable on the web. IMHO it's a much more pleasant language to work with.

Funnily enough I am doing the exact same thing. In my mind Golang has a temporary spot on the roster, and I'm actively looking at Rust, Nim, and a few other languages for writing some server projects in the next few years.

If Google isn't careful, Microsoft is going to cannibalize their enterprise space. I think Microsoft has been getting the same vibes and it really seems like they are finally trying to be team players and not just catering to the low-hanging fruit of non-tech savvy folk.

I wouldn't be surprised if they take the gamble on releasing a (free?) online enterprise office application suite as some sort of MS Office Lite. Google is no longer the hero of innovation we all thought they once were, and slowly but surely, smaller niche companies and serious tech giants with the singular goal of squashing Google will replace their core products with products that are more open source, privacy-focused, and with real customer support.

I remember when I wanted to work for Google. God, I remember when I wanted to work for Facebook. What the fish was wrong with me.

Microsoft already has released Office Online. It's lite versions of Word, Excel, and PowerPoint. They're quite good. They get certain enhancements when you buy an Office 365 subscription (in addition to the full applications).

IMHO there's nothing wrong with wanting to work for Google/Facebook/any of the big 4/5. Get one of those on your resume and you are set for life, even if the work isn't all it's cracked up to be.

I didn't know about Office Online. Glad to hear Microsoft is doing that.

The problem with Google and Facebook is that I've come to realize our ethics are extremely misaligned.

I'm not interested in building exploitable systems filled with damaging metadata that can be abused by totalitarian states or vengeful employees.

I'm not interested in forcing Africa into a world where Facebook and Google are synonymous with the internet.

I'm not interested in running an echo chamber that 1/4 of the world's population uses, subjecting others to political influence and carefully syndicated information that influences their thoughts and behavior, while providing massive quantities of foreign user data to three-letter agencies interested in squashing social uprisings in third-world countries.

And I'm not interested in shutting down my moral compass by accepting corporate propaganda that tells me what I am doing is right and for the benefit of mankind.

Only reason I ever wanted to work for either was the prestige, the badass lunch buffets, and the feeling that I'm doing something important, positive and ground-breaking. But those are all selfish desires and the last one isn't true.

Many of those reasons you wanted to work for them are very subjective. While you mentioned that the last one is not true, I find that the Google Deep Mind project to build a Starcraft AI is extremely interesting.

I'll agree, they are in some ways subjective, but I have to work with the assumptions and motivations I have when it comes to choosing my employer.

Now Deep Mind... I would work for Deep Mind. Or even just intern. I would love to work with some of the ML pioneers even if just for a few months and just soak in everything I can.

On the contrary. I've spent much of my career in the conservative "enterprise" world... and outside of the HN echo chamber, most established companies aren't yet making heavy use of Google products at all. Largely due to the cultural disconnect being discussed here.

Sorry, but "We'll give you 12 months notice!" is still nowhere near "enterprisey". Try a decade.

>Try a decade.

And even longer if you're willing to pay outrageous sums of money.

I work in the conservatibe enterprise world, and vendors dropping services with much less than a decades notice -- usually, but not always, with a recommended replacement that they are also willing to sell you -- is quite common.

I think Google's problem here, inasmuch as it has one, is marketing and relationships, not substantive policy.

First, isn't the HN echo chamber actually an important constituency for Google if they want to sell to businesses? Second, I think once you get one of these "over my dead body" reactions going, it has to hurt business. Consider the lengths Microsoft has had to go to to try to overcome their negative reputation in the tech community.

Except Site Search is primarily used by non-technical people. It's widely used on forums and common on WordPress sites.

Depends on the field – e.g. in the cultural heritage space, many people used Google Reader and that still comes up any time someone mentions a Google service: “What will we do when they cancel this?” That doesn't mean that people won't use something but it definitely encourages a healthy discussion about lock-in.

I don't care as long as "site:example.org query" still works on google.com

Effectively does the same thing.

I know a lot of big company CTOs –– for old, boring companies like insurance (blah!) and real estate (yuck!) –– and they are very concerned about tech companies' willingness to keep services around for the very long haul.

There's more than a year of notice, and there's still a solution that provides substantially the same service. Long term can't mean forever, especially for services that can't support themselves.

Their search appliance (GSA) is being replaced with a cloud search product. I would guess once that is released it will make the GSS product a bit redundant, so it makes sense to discontinue, the timing is a bit strange though...

Note that Google custom site search is still available: https://cse.google.com/cse

The main difference seems to be that enterprise has custom branding and is ad-free.

It is, but they're in the process of deprecating and removing its most powerful features, such as the "linked CSE", where you can specify the URL of an XML specification file that describes the domains and paths you want to search. The huge advantage of a linked CSE over their 'hosted' alternative, aside from convenience, is that you can generate this specification dynamically based on a user query. They also recently made a policy change disallowing multiple CSE queries resulting from the same user's search. (IE if you want to give the user search results with several different filter parameters.) Both of these changes had a fairly significant, although fortunately not critical, impact on our business. It really looks like the philosophy is shifting away from these sorts of flexible tools.

On the other hand, Bing recently released a completely updated version of their paid search API. So on the plus side it looks like they aren't ready to give up on custom search yet. On the downside, they've tripled their prices.

Since they pushed Excite out of this business, who will take up the slack?

Stucked in time: http://www.excite.com/

look at the source. Something makes me think this will work in Netscape 4.

edit: nope. http://imgur.com/a/Jt38q not really.

edit 2: Here's a HP B-series PA-RISC from my closet running actual vintage NS-4, also no, http://i.imgur.com/tt7AYqj.png (the top window is a search for the term "cat")

Google however, almost still works http://i.imgur.com/kdTd1AU.png ... I give them an A. hn and kernel.org get a security algorithm error, reddit gets an i/o error as does craigslist and yahoo. wiki.c2.com has a perpetual spinner. netbsd here makes me sad to be such a fanboy: http://i.imgur.com/RxedDF6.png

http://gnu.org is totally acceptable, even in Netscape 3! http://i.imgur.com/RImHQjM.png

http://fsf.org errors at an appropriate place: http://i.imgur.com/g8TF8Uc.png but after that is usable.

http://berkshirehathaway.com loads perfectly and looks the same.

c2 recently became obsessed with "federated wikis" apparently due to some unimpressed person threatening to batch-delete the entire site. Unfortunately JavaScript-based solutions became interesting.

I'm mildly curious what would happen if you tried to build NetSurf on that thing. It compiles for AmigaOS and RISC OS...

First let me apologize to the HPUX wizard that will very likely read this for being such a bumbling amateur. Anyway, here I go:

H3:54/raid/hp/netsurf-all-3.6$ gmake --version

GNU Make 3.80

H3:55/raid/hp/netsurf-all-3.6$ /usr/local/bin/gcc --version

gcc (GCC) 3.3.1

H3:55/raid/hp/netsurf-all-3.6$ gmake BUILD_CC=/usr/local/bin/gcc CC=/usr/local/bin/gcc


gmake[1]: Entering directory `/raid/hp/netsurf-all-3.6/libwapcaplet'

/raid/hp/netsurf-all-3.6/inst-gtk/share/netsurf-buildsystem /makefiles/Makefile.tools:403: /Makefile.gcc: No such file or directory

/raid/hp/netsurf-all-3.6/inst-gtk/share/netsurf-buildsystem /makefiles/Makefile.tools:460: /Makefile.pkgconfig: No such file or directory

Makefile:40: /Makefile.top: No such file or directory

gmake[1]: * No rule to make target `/Makefile.top'. Stop.

gmake[1]: Leaving directory `/raid/hp/netsurf-all-3.6/libwapcaplet'

gmake: * [/raid/hp/netsurf-all-3.6/inst-gtk/build-stamp]

Error 2


Looks like some serious effort. To put things in context, I can't do remote X to a modern machine as in

$ DISPLAY=desktop:0 xterm&

This gives a protocol error. So instead I'm running it all through a vncserver, which uses a more legacy protocol. I've tried things like

$ Xnest -query hp :1

but I get the classic CDE hour glass, a black screen, and nothing more.

It's honestly nice to pull this thing out of the closet just to remember how unfriendly things used to be. You type a command and realize "well gee, this thing doesn't have that. Alright, here's a more painful way..." You don't even get things like arrow keys and backspace for free. Gotta stty them.

I must admit that I would _really_ love to have a go at this HP-UX box. I have some ideas about how to fix the X issues you mentioned, and getting remote CDE working to your normal X display would be awesome too.

To be completely fair, building NetSurf may simply not be possible with the toolchain the machine has - I was maybe-1/3-joking about trying it :P - but it would be an awesome challenge to see if it could be done. I suspect it may just be possible.

I emailed you.

"Looks like some serious effort. To put things in context, I can't do remote X to a modern machine as in

    "$ DISPLAY=desktop:0 xterm&
"This gives a protocol error."

Ok, I'm impressed. That's the first time I've seen that.

I believe it's relatively recent. I've heard second-hand from better men than me that they are dropping a lot of legacy things. As far as I can tell, the HP is running X11R5. It doesn't do -version, the man page is for R5 and the binary is dated "Oct 27 1997" ... so yeah, 20 years ago. It's honestly a reasonable thing to break.

X11R5 is older than I am (and I'm old). Wow.

That would explain it.

Seeing Macromedia's logo really hit me in the nostalgia.

I remember tinkering with Flash back in the day, even before it was Adobe.

I'm not sure this entire stack has been updated since the early 2000s. No SSL/TLS, and check the HTTP headers.

Wow, that was a thorough testing.

Holy crap, I worked on parts of the UI on that site. No idea it was still online.

Really? What years?

I guess it'll stay up as long as it generates revenue via ads. Its still someone's homepage I suspect. The elderly still using AOL perhaps?

I've never wanted to see a revenue graph more than that of Excite. I would suspect there would be a strong correlation with AOL dial-up subscription rates (perhaps causation too).

And if that's not good enough... they have a LITE VERSION. http://lite.excite.com/

Simply making a version of X that loads on dialup is itself a market segment. Clone some popular websites but sacrafice everything.

Also I've long wanted an HTTP caching mechanism that permits cross origin caching, so people aren't redownloading the same dependencies. Essentially the server responds with a digest and if the browser has a match from any domain, it forgoes the download.

Is there a hard to trigger collision attack possible? Sure but I don't care.

Yes, but Excite Lite no longer supports the webmail.

My login/password still works and custom theme is there !

Mine too, as I found out two weeks ago. Excite is back to my primary webmail.

and yet it has a ranking of 8,689 with huge traffic uptick since Jan 17 as per Alexa today .... does that mean people still use it regularly?

Check the ' Top Concert Tickets' section and performers. Seems mainly for people wanting to stay in a former era, which is not bad as they get something they find useful. Also a fair bit about Home Security, Medical and Nursing, Online Schools and Criminal Justice. I think they have an audience.

Very much so. In a world where Google is watching everyone and Yahoo is getting relentlessly attacked, why not go back to the old standby?

Looks like there are doing some 1990's SEO too:

Home Security Systems Arizona

Home Security Systems Atlanta

Home Security Systems Augusta

Home Security Systems Austin

Home Security Systems California

Home Security Systems Cleveland Ohio

Home Security Systems Columbus Ohio

Home Security Systems Dallas...

Crikey, I just saw a DeLorean fly by.

Wow. This page is actually strangely appealing. It loads fast. I don't have to click off to another page to get the news headlines. Ads are restrained to just a couple square boxes.

Amazing. It even renders more or less decently on mobile.

I'm going to start a new trend in web design. I'll call it 56k design.

> I'm going to start a new trend in web design. I'll call it 56k design.

That was pretty much my criteria when I was making forum software: https://www.lfgss.com/

It's still too heavy and slow, and there's still some things I could do better. But that's nitpicking, it's pretty good and when I finish re-structuring all of the code it will be a single binary install for those who want to use it, or a split-binary Web-UI + API for those who want to only do certain parts (i.e. customise and host the front-end and not care about the back-end, or to scale the background for heavy mobile use without that being serving front-end traffic).

My goals for that software in the next few months:

1. Same-origin everything

2. Eradicate as much JavaScript as possible (graceful degradation to the extreme)

3. What JavaScript cannot be eradicated, do natively (not using jQuery, etc)

4. Single codebase, single binary install from a single `go install` command

We've built the Thredded forums engine (https://thredded.org) with the same goals in mind. The entire CSS is 10KiB, JavaScript loads asynchronously and is optional.

That's delightfully nice to use. Damn good job.

Oh hey, I love the lfgss site design. It just gets shit done and looks decent on everything I own.

Oh, people have already tried this. Very recently I think someone posted their website on here saying it was amazingly fast etc. Anyone who used the web back in the 90s just remembered it as how things used to be. A Google search was virtually instant over a 56k modem. Now it's noticeably laggy on a broadband connection. If you used Firefox then web pages started rendering the instant enough HTML had been downloaded. None of this waiting for the entire thing and then running over it all with Javascript to generate the "real" render.

Well as long as you still transpile 10MB of es7 to show a paragraph of text and an image, we're all good.

It's almost like a knock off of a Japanese "portal" except its not but rather its sort of a fossilized 1999, except it has SSL.

And to think they had the opportunity to buy Google for 750K after Vinod talked Page and Brin down from 1MM

It may be a couple of boxes but that link spam is something else, thought I was a domain parked page for a second there.

But that like half a page of link spam loads faster than a single banner ad on most sites.

From the top of my head, I'd say this is a great opportunity for Algolia.

Google owns this and they are giving it up, why? My guess is they dont make enough money this way (with search as a service). That is probably just a long term concern but an important one.

Yep, that's a good opportunity for them, for sure. I'd start calling up universities and get a foot in for exposure, list them as enterprise clients, etc.

Coudlnt you just hid their ads with css ?

No doubt that would be against their ToS and they would stop serving you.

Besides, hiding ads is not the same as preventing ads from being loaded.

Enterprise and government customers are often concerned about data leaking out through advertisement networks, and hosted ad services are not uncommonly used to inject malware into specific sites. Hiding ads using CSS would not prevent these problems.

this is why i'm not completely 100% with google photos. i'm ready to spend the monthly fees to store all my pictures in its original resolution, use google photos to manage/search my photos. except i don't want to commit to using it and then finding out google doesn't want to support it and just dumps it.

same thing with google voice. i use it but i've never committed to using it as my main phone number.

maybe it's a catch-22, but maybe google should stop sunsetting so many products so as to cause people to have no faith in their products. pretty much the only untouchable thing is gmail since the negative press for that would probably end google to the general public.

Gmail and Google Drive, my biggest ties. I can't see either of these going away, because like many mentioned these services are consumer based which Google seems to care about more. Well, I heavily use Adwords professionally and that support used to be pretty decent - though I feel it's gone downhill very quickly lately. Someone will answer your call but the expertise is lacking which is fine for your level 1 support but even their offshore "specialists" only seem to give cookie cutter optimization advice or the famous "add more money."

They've recently transitioned to this "beta" health checkup feature that gives the same advice except without people! It's automated and to be honest, it's even better than the support I was getting. This could view my stuff and if I was missing a critical component (I have a lot of ads running) it'll call it out.

I don't know what happened to Adwords but about a year ago, I had amazing onshore support and the person listened to business case scenarios and we would have discussions on possible strategies and ways to save money but be more effective.

I would love to be a fly on the wall on some of these higher up Google exec meetings on where they switch priority on a dime.

The strategy makes sense to me. Hire people who deeply understand the product to support users. Get as much feedback and use cases as you can. Then eliminate support costs once you've found your place in the market by first replacing staff with lower paid, less knowledgeable people, then automate it.

I'm right there with you. Long term Google product user here (Android, Fi, Gmail, Drive, Photos, Etc) and I'm winding it all down, except gmail which will remain my spam-mailbox.

Gmail is shit. Gmail is Email Extraordinaire but if something goes wrong and you lose access, good luck gaining it back. Try to call somebody. This, and all the privacy issues.

Account recovery is a shitshow everywhere, either it's almost useless or it's a huge vulnerability with no good middle ground I've seen.

For a moment I thought this would be a good use case for federated authentication, but who would you even trust to run that securely?

How about a company that you pay and that you can call? (and maybe submit a passport copy if need be or something)

They sell email for your own domain (you can also buy a domain there): https://www.infomaniak.com/en

Bonus: Based in Switzerland, no EU or US jurisdiction.

I'm betting that Google Photos will necessarily stick around in order to keep Android competitive with iOS (and perhaps Windows.) I desperately hope they make it a paid service, if needed, rather than shutter it.

I've fully committed and have uploaded 40,932 photos. All of the automation has been amazing.

I'm not trying to be a dick and am genuinely curious. What are you going to do with those 41 thousand photos? Do you ever go back and look at them? I see lots of people taking a ton of photos, they get stored in the cloud automatically, and then... what? Does anybody ever go back and look at that bad photo of a fish they took at an aquarium five years ago?

This may sound defensive, but its just a genuine answer, as I agree that many photos are pretty pointless.

In my case I uploaded every photo I've taken since I got my first digital camera in 2000, and I do look at some of them.

To try to answer your question directly: I just checked and I have roughly 100 albums with roughly 50 photos, so I expect that I'll go back to those when I want a trip down memory lane, etc.

As for the other ~36k, I don't have any statistics but:

* On ~10 occasions I've sent emails to friends after getting one of the "this day in YYYY" notifications with a link to the photos

* On a few (~4) occasions, I've used the faces grouping to find old, fun photos of friends

* Yesterday I happened to get a "This day in YYYY" notification that reminded me about an event I photographed at my neighbors house, but had forgotten to send the photos (7 years ago, oops)

* I'd rediscovered some DSLR photos I probably would have missed thanks to the "approx location tagging" (that taps in my phones location history)

* I share photos far more frequently since its free, unlimited and auto-upload

* The auto-animations and auto-panoramas have made me go back and rediscover photos from some old trips

* Just the other day my Mom was interested in a photo she remembered me taking on a trip in 2013. I was able to search the place name and find the photo in seconds

I started down the path of Google Photos because my old workflow of manually merging smartphone and DSLR pics on disk and then not having a good centralized album management system felt clunky.

I'm surprised by how often I find myself going to http://photos.google.com now.

The value of Google's automated processing of images can't be understated. Last year I uploaded roughly 50,000 photos to Google, which was every photo I had taken since 2006. Google automatically created albums for every vacation I'd gone on in that time, plus many of the major events. Weddings, concerts, etc, were all identified.

Also the face search is probably the most brilliant piece of it. I began meticulously tagging photos in Lightroom a few years ago, and had probably 15-20% of my photos tagged with location and people in the photo. It was a pain in the ass and took forever to get to that point.

Meanwhile it took less than a week to upload everything to Google Photos, and now I can search through over 10 years of photos by person. You can also search by event, object, location, etc ("coachella," "sailboat," or "haiti" as examples).

I completely agree. In other software labeling (albums, location, faces, tags, etc) is tedious, time consuming and fragile. In Google Photos I barely even think about it.

Do you have a local backup of those photos? Every now and then someone's Google Account gets mistakenly shut down by, I presume, some automated risk assessment process and they can have a really tough time contacting anybody to rectify the issue. I'd say you can use Google's services as tools, but not as the basket where you put all your eggs.

Dataloss/lockout is a great concern of mine and I am somewhat exposed. For DSLR photos, I persist them on disk, then upload to Google Photos. For smartphone, I periodically copy the new ones to disk. Then that disk is remotely backed up. However, I am exposed to losing all of my GPhotos meta data (primarily albums and captions) as well was the smartphone photos since my last manual backup.

I have accepted this tradeoff. It would be pretty devastating to loose all that organization, but its recreatable and the value I get out of Google Photos trumps that risk.

As an example: I wish so dearly I could have more than a few physical photos of the summer I spent in England 12 years ago. Of the friends I made, the things we did, and the places we visited.

As you say, I barely look at my pictures from 1 or 2 years ago (except for the "this day X years ago" feature, which is awesome). But I'm positive that 10 years from now I'll be glad to have them, for the nostalgia trips.

I have uploaded O(10K) photos to Google photos, with O(1K) albums. I find it immensely valuable. Here are some scenarios where I actually use the uploaded pics:

- family reunion: watching videos of family members taken over the years (nice integration with chromecast to watch on tv)

- meeting friends after a long time: quickly retrieving super old pics (good search abilities with album names and tags/captions)

- casual discussions with friends/family about a specific experience (eg. trips to a particular destination): quickly retrieving pics taken there

Coupled with the fact that there is unlimited storage, ability to share across as many channels as I want (email, FB, WhatsApp...), seamless backup and sync across my various devices, ability to create shared albums with other people - I would say this is one of the rare great products Google has released in the last ~5 years.

p.s. fwiw, I also have a local backup copy in case Google messes with my account.

I find a nice way to be reminded of some memories is to have random pictures show up as my desktop or phone background once in a while.

I also go back through them and pull out some for use during big lifetime events (10*n birthdays, weddings, retirement, etc).

This is the reason I pretty much stopped taking photos altogether. I realized I almost never go back and look at them.

Photos are like Steam games, maybe when you are in the hospital for a stay you can knock both out in one go, eh?

> I'm betting that Google Photos will necessarily stick around in order to keep Android competitive with iOS (and perhaps Windows.) I desperately hope they make it a paid service, if needed, rather than shutter it.

Photos is a paid service already, with a free tier.

With similar concerns, I wrote photobak so I always have my photos locally: https://github.com/mholt/photobak

Why not just use CaptureOne/LightRoom, and have encrypted offsite backups? Using a specific product to "store" pictures sounds silly.

I can't speak for the person you replied to, but I rely on a product to "store" pictures so I can access them when I'm not at my main computer. I'd like to show them to others using my phone, share them with other computers (like my wife's) so she can edit some, or add to the collections, and they are also accessible to the machines hosting my web sites, so that others can browse the ones I choose to publish.

Encrypted backups are great for recovery, but they don't address many other useful features of these additional products.

I use Lightroom, but it's fairly "dumb" in its approach to photo management: no features, no surfacing of old content, just a useless grid. GPhotos on the other hand surfaces old photos in interesting ways, shows you moments from the past and even sometimes creates cool edits on its own. It's actually pretty magic, and feels years ahead of Lightroom.

Well, they're also for different use cases, so I don't think "ahead" is the right word. Lightroom is pro-oriented and assumes that you know what photo you want to find, whereas consumer photo apps are more about browsing and availability than anything else (hence the autogenerated montages and cloud storage integration). But that is exactly why I use a combination of Lightroom and Apple Photos. (edit: Though I wouldn't mind having some of these slick organization/discovery features in Lightroom as long as they didn't ruin the rest of the experience.)

What would happen if they wanted to sunset Voice? There's a large bank of phone numbers people migrated in there. I'm guessing they could spin out Grand Central again as its own business?

Or they could just say, migrate your phone number elsewhere.

They might not kill Gmail but I wonder about Inbox. Is that receiving much love?

www.famipix.com - online since 2005, and free for schools

It's not ad-supported, so Google has to actually sell the service and collect money from large numbers of small users. Google is not good at that.

Cost of sales / cost of revenues is in general a huge problem, and there's a long history of benefits accruing to businesses which work out how to bypass that.

Magazine subscriptions (bundling), utility services (a novel concept in the late 19th century), governments (taxes), banks, etc.

It's why the enterprise market is far more appealing, generally: make a few large-ticket enterprise sales (mostly to the F-10 - F-500), even if they do take forever to land, and roll in the recurring revenues.

Retail sales is far more about the distribution deals than the register sales (the second follows the first). Competition is generally over the marketing channel and/or real estate (look up "dark grocery stores" for more on that last).

Or as I'd realised the first time I heard Milt F. rattle on about TANSTAFL, there's no such thing as a free lunch revenue-capture system either.

Google are only interested in selling things by the billion or billion, therefore things like the robots have to go. This is one of those products that does not sell to millions and therefore it has to be got rid of.

Not sure when the 'must sell billions' idea came along, it was before the Pixel phone but is a relatively new idea for Google.

It's sort of the startup philosophy in reverse. "Only do things that have economies of scale." As if Google is just one person controlling a billion robots, and can do anything a billion robots can do, but can't do anything that requires more than the oversight of a single human for the entire project. (In est, Google is the player-character in Factorio.)

This is an argument for automating top management. There are companies that choke on limited CEO bandwidth, and Google seems to be one of them. Apple under Jobs kept the product line so small that one person could manage it, which is one way to do it. Google, though, keeps trying to broaden their revenue stream beyond ads, without much success.

This is a classic management problem. The first person to really get a handle on it was Alfred P. Sloane, who headed GM and wrote "The Concept of the Corporation". He figured out how to organize a very large company effectively, and GM's management structure worked well for half a century. They beat out everybody else until Japanese competition appeared. (GM probably stayed with Sloan's structure too long, but nobody else really knew what to do.)

There's the argument that, rather than one company having a broad range of products, there should be many companies, each with a narrow product range. It's cheaper to run a fund than manage a conglomerate. The conglomerate concept was that "synergy" between the business units would reduce cost. In practice, this rarely works.

There are some successful companies with very broad product lines. Mitsubishi and Samsung are two of the biggest. General Electric was the classic US example, but they've backed off from most of their consumer products other than major appliances.

So basically, the Google-character has too low APM. Maybe time to hire StarCraft progamers for executive positions?

The similarity between the words "progamers" and "programmers" was really irritating to me in that line. :)

They manage that with Google Music, recurring Play store payments etc.

With a couple orders of magnitude more users, yes. Willing to pay $100/year to remove branding on a search engine you add to your website is far, far less common than "willing to pay to listen to music".

Plenty of small businesses spring for gmail and Google Apps for Business

Which are also a couple orders of magnitude fewer in number than music-listeners.

Nonzero to be sure, but nothing like the numbers they're used to in free, ad-supported products.

Given the big mess they had when Google Drive migrated to Google Play subscriptions I wouldn't exactly say that they're good at it. http://imgur.com/a/gzkZ5

and G Suite (formally Google Apps for Business)

Uh, CSE is stayin and you can pay to remove ads from CSE. Once you run out of searches they send you a link and you click on it and press 'ok' under business. It uses your stored business card. Its pretty quick and I can't imagine it being an issue for anyone. Google also takes payments for other services like Play Music, Youtube Red, Apps, etc.

I imagine the issue is that GSS hides a lot, (all?) of Google's branding while CSE doesnt as its a page hosted by Google.

I think by paying to remove the ads from CSE you're signing up for GSS. If you sign up from the link it takes you right to CSE.

This actually is making me fairly frustrating. Me and my team were literally just finishing the GSS integration and all of the sudden the only option google is suggesting is "will automatically convert to Custom Search Engine (CSE). Custom Search Engine is an ads-supported product "

Per the Google's statement: "On April 1, 2017, Google will discontinue sales of the Google Site Search. All new purchases and renewals must take place before this date. The product will be completely shut down by April 1, 2018."

If I were you and you really had already finished the Google Search dev work you referenced, I would pay for service as soon as possible for the duration up until April 1, 2018. Use the service, see how user users are really using it, then use what you've learned to plan for the future.

Thank a lot for your suggestion :) We will do that as soon as possible.

This could be a good opportunity for DuckDuckGo to make some money and improve awareness of their brand.

No kidding! I would totally give such a service a serious look for future client work.

It's happening right now: https://duckduckgo.com/search_box

That's not quite what I'm looking for - I need an API. I need to maintain client branding on site without mentioning the search service provider - and I would expect to be billed for that.

The conversion here at HN on this topic has been useful in pointing out such services that I didn't know about, for which I am thankful!

Differences between GSS and CSE: https://support.google.com/customsearch/answer/4541888?hl=en

As someone who just finished moving a custom search result site implementation from GSA (discontinued) to GSS this is highly outrageous. No access to the XML/JSON API is a showstopper.

Last couple of years or so have been an eye opener for me on how I choose libraries, tools and services for my project. One important lesson was that never choose a service which don't have plug and play open source alternative. Even hyped start ups go out of business and big companies like Facebook and Google shut down services.

Third party APIs are a even bigger risk. Facebook and Twitter APIs used to change quite often 2-3 years ago. (Haven't personally worked on this off late though.) APIs and services which seem too cheap are generally the ones which should be avoided. Because it is more likely that the companies running these will not find a sustainable business model.

Even Open Source projects get abandoned quite often. Sometimes a team puts together an impressive open source project and then they get acquired by a big company which results in abandonment of the project. I have one client who has his startup riding on Kurento Media Server for WebRTC calls. This product is in development from last one and half years. Now, after Kurento's acquisition by Twilio, Kurento's future is not clear. Also it's not a project which any body can just get into and start developing it further. Thankfully Kurento is not abandoned yet by the development team, but I had read somewhere that they are looking for new maintainers.

The lesson learned is only depend on mature open source projects and depend on PaaS services which make those open source projects available as easy to use services. Even if that PaaS service go down, there is a high chance that you will find someone who can deploy and support it for you.

However, if you are depending on a third party API or a relatively unknown Open Source Project for critical functionality, treat it as a big risk in your product's future road map and be prepared with a contingency plan.

Right now in cloud services there are way too many products which are not easily replaceable. But these are easy to use and sometimes solve really difficult problems. So the temptation to use them is too high. So they still end up in the stack, especially when the clients themselves push for these services to be used.

Of course, another product shut down by Google.

Custom search has large maintenance costs--a lot of effort which really isn't interesting for any developer to work on. And apparently companies aren't willing to pay enough to overcome that. I don't think Google has any obligation not to shut down products.

They have no obligation, but it hurts adoption of their other products since it has happened often.

They should be mindful of that trend since they already have gotten a bad rep because of it. At some point in the future, when they really need it, few people will jump on their new band wagon.

And we have no obligation to expect their next product to stay alive long enough for us to use it properly.

On the German page I get when following this link, there is still no shutdown notice visible but only the "normal" marketing page for Site Search (+pricing). Does the shutdown only affect US customers or does Google fail to communicate this internationally?

Hum, I'm seeing the notice in the UK.

I suggest using Coveo, it's the recommended thing to replace google enterprise search. Much more intelligent and better integration with the data.

It's already used by majors companies like GoPro, Adobe and L'Oréal.

Yes I worked there once :)

Another vote for Coveo (as someone who still works there)! The overall integration, security features, analytics, connectors and machine learning functionalities are beyond anything that Google or pretty much anyone else has to offer.

FYI We're also hiring and growing very rapidly! careers.coveo.com

Coveo does look cool but at my level, $2k/yr, it's far more expensive!

Detailed article with comments from a Google spokesperson:


What are some competitors to site search?

Algolia, SwiftType, sooqr.com, Hawk Search (for ecommerce stores), Inbenta.

HN and ProductHunt runs on Algolia.

Edit - some more details here: https://siftery.com/categories/api-and-developer-tools/searc...

Cludo is another competitor. Disclaimer I am the CEO there. We have previously gotten quite a few customers to switch from GSS.

I've used and am impressed with Sajari: https://www.sajari.com/

You can try searchIQ. I am using it for a while and my website search looks good so far. Though i am still using free version but they have also released Pro version with added functionality. You can have a look at that also.

We've been using https://swiftype.com/site-search for a couple of years. It's great.

Pricing starts at $299/mo. Bit steep if you're a smaller company.

They dropped the free plan some time ago, although if using their WP plugin on a wordpress site, you can still make use of it. I also found SearchIQ but never used it to be honest, so I can't really tell. Algolia is quite powerful too.

Postgresql's tsvectors and GIN, Elasticsearch.

That's on a different layer, though. If you already have your raw content in a database somewhere, you can push that into Elastic or Postgres without too much effort, but if you need to parse your HTML documents first, good luck with that.

Azure Search (Disclaimer: I am a MSFT employee who works on Azure Search)

So, more room for Algolia et al?

Here's a quick and easy search form for your site:

    <form id="searchform" name="searchform" action="https://google.com/search">
      <input class="searchtext" name="q" type="text" value="" placeholder="" maxlength="300" />
      <input name="q" type="hidden" value="site:www.yoursite.tld" />
      <input class="searchbutton" name="submit" type="submit" value="Search" />

Where is the shut-down message? I just get the pricing page for GSS in Dutch.

EDIT: I was also able to create a new site search engine.

> Where is the shut-down message?

Try English version of that page: https://enterprise.google.com/search/products/gss.html?hl=en

> EDIT: I was also able to create a new site search engine.

The shut-down message says "On April 1, 2017, Google will discontinue sales of the Google Site Search".

Amusing how they don't feel the need to inform non-english people.

That message is possibly in a to-be-translated queue.

For something like this.... maybe not all languages, but just the ones that analytics detect significant traffic from.

I checked Dutch, German and French. Nothing.

I see it on the English site. I bet if you set your language to English, it would show up.

I have only English in Accept-Language, but Google still displays the page in Polish. Yay. :-|

I've always wondered why Google doesn't just slowly raise the prices on things like this until either everyone switches due to attrition and they can shut it down noiselessly, or it becomes profitable.

There's been a few services (Google Reader and CodeSearch) that I would have paid a couple bucks a month for.

Shit, I use this for image search. Basically we let users submit the title of items they own (for everydaycarry.com), we do a google image search for that title and display back the top 5 results from googles image search to the user, letting them click on the appropriate product. We then link that product to the google image source URL.

Anyone know of a alternative product that would let this keep happening? When we implemented this originally, I did not find any competing options from either Microsoft, Amazon or Yahoo- That was 2 years ago though, so things may have changed for image search API's.

We are looking for links to products on other sites- for instance https://www.google.com/search?q=gerber+shard&source=lnms&tbm...

Also, the link you give above is just googles normal link. You cant programmatically call it, or google will start throwing up a captcha which will stop automated queries.

I have a feeling this is geared towards pushing fortune/alexa 1000 into buying/using the google enterprise search appliance, which they charge a pretty penny for but don't allow you to see inside of (black box on the network). I was looking at one and got quoted 50k, and then when I balked and stopped communications with the sales guys they started offering price reductions... like 30k for a blackbox search appliance on my network is really something enticing!

Unlikely, the GSA product is also discontinued. This is a push to a google cloud alternative i would guess.

That's why I think twice before embracing some Google Product.

What is so hard technically about site search that there doesn't exist open-source solutions for people to run on their own sites or rent for a small fee?


Also posted changes coming to the CSE API.

Pay per DNS lookup? There's a SaaS somewhere in there.

More like IaaS

Any site can free themselves from the Borg: https://duckduckgo.com/search_box

They shut down Google Mini few years ago. I never really understood why would they do that. Looks like they are on the route to shut down all enterprise services.

How does it relate to google search console ? I don't understand the relationship or the difference between both (unless there are the same ?).

Google Search Console is Google Webmasters Tool, it's a different thing.

That's a pitty for people like me who are using it.

Although you have to say the product had its flaws on larger scales.

If you are running an eCommerce site and you are looking to improve your search www.prefixbox.com can be an alternative. We are developing it for 2.5 years now and it is a data driven solution.

BR, Istvan Simon

Aprils fool 2 years in advance.

April fool?

wow - shocking;

I wonder if anyone will move into this space instead. hoping for duck duck go or a new startup.

What? Why? You can still do the same search by typing site:URL in Google so it's literally zero investment by them to keep it workinf. And these people were paying for it.

I can't help think this has something to do with their AMP articles. They don't want you to leave google, ever

I believe that this Enterprise Search can be used for pages that are not freely available on the Internet. I've seen it used on portals where you have to be logged in to access the content.

Not sure how much it overlaps with https://enterprise.google.com/search/products/gsa.html or what the future for GSA is, given Google's track record for discontinuing products.

I doubt it has anything to do with them not wanting you to leave google, more that there isn't anything in it for them.

I suspect they thought at one point that this would be a freemium product and people would upgrade to the white-label version, but maybe they weren't, or not enough.

Now, if somebody wants google on their site, they have to pay google, or force visitors to go to google.com, where google will show them ads.

Paying gets you white-label branding and no ads.

You can hide their ads with css. Probably against their tos, but I haven't been busted for doing it in 2yrs.

Why? According to pricing page at least 100 USD a year for basic product, that's why.

I read this and assumed it was just an early April Fools Gag.

Actually, I'm still not convinced that it isn't linked to April Fools.

I also think this is some kind-of April Fool Joke

For a brief moment I've read "Google shutting 'Down Side Search'".

LOL -4 karma on that one.

Applications are open for YC Summer 2023

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact