Ask HN: Google employees do you have a better version of Google internally?
That actually shows meaningful results, git repos and other information; or do you have to use the same trash we do?
7 comments
Internal company documentation and search is an interesting problem that I have been thinking a lot about.
One big issue is that everyone knows its an issue, so people add more places to store information, which inevitably becomes another place to search which makes it hidden, and then as things move there are inevitable dead links all over the place. We need an intranet to be an actual intranet with all information there, and properly indexed.
The biggest "solutions" here are also some of the worst offenders:
1. Google Drive / Microsoft Network Share / Drop Box Etc. These are filled with documents, some companies have more than one of these. These are places where information goes to die, and lives entirely in people's heads, so you have to talk to someone, have them find it, send you a link etc. If there is a search, its for the names and not the content. They are rarely in a good format or place to easily do a find | grep search.
2. Confluence is awful. Confluence is directly backwards, its information at the team level, instead of at the topic level.
3. Sharepoint purely awful. Its searched poorly, barely works.
4. Wikis, very few places uses these for some reason in favor of worse. A single source of truth wiki I think is the optimal place. Or if they do its like a MS Team Wiki or a Gitlab wiki that is broken off of all other places and not indexed so its disconnected.
One big issue is that everyone knows its an issue, so people add more places to store information, which inevitably becomes another place to search which makes it hidden, and then as things move there are inevitable dead links all over the place. We need an intranet to be an actual intranet with all information there, and properly indexed.
The biggest "solutions" here are also some of the worst offenders:
1. Google Drive / Microsoft Network Share / Drop Box Etc. These are filled with documents, some companies have more than one of these. These are places where information goes to die, and lives entirely in people's heads, so you have to talk to someone, have them find it, send you a link etc. If there is a search, its for the names and not the content. They are rarely in a good format or place to easily do a find | grep search.
2. Confluence is awful. Confluence is directly backwards, its information at the team level, instead of at the topic level.
3. Sharepoint purely awful. Its searched poorly, barely works.
4. Wikis, very few places uses these for some reason in favor of worse. A single source of truth wiki I think is the optimal place. Or if they do its like a MS Team Wiki or a Gitlab wiki that is broken off of all other places and not indexed so its disconnected.
Confluence also has a default-deny policy on editing, which I think is crazy especially in an internal enterprise context. If one of your employees is vandalizing wiki pages then you can see their name in the change history and fire them. Prior restraint makes no sense.
I'm not sure if it's the default but I know it can be change. If your company has chosen to go that route, that's on the company, not on Confluence. Anyone can edit Confluence pages at my company.
Similar defenses are made of JIRA. The thing is: the whole reason these products exist as distinct from their open source or practitioner-favored alternatives is their enterprise feature set, which is to say their affordances for being locked down and micromanaged. Buying Confluence and then running it wide open is a weird choice.
To be honest, Drive does a good job at indexing also the content of the documents. My gripe with it is that when I search something I get all the results of things that people shared with me directly or indirectly though groups. If I could have the option of searching only things that I actually accessed, or even edited/commented I would be really happy about it.
I actually seem to have the opposite result. I seem to only be able to find documents that were shared with me, or that I accessed myself. I want to be able to find documents that I have not actually seen, yet have access to view. Confluence is a bit like this - i can search all of the spaces and get results from other parts of the org. but drive seems to push results to the top which are more localized to my immediate previous use
I always get lost in Google Drive. I don't even know what I am looking at in Google Drive. I wouldn't say, one of the worst UIs in existence, but it is hard to give any more credibility to it. After years of use, I still don't know what I own, what is shared with me, what others own, how are docs organized, etc. At least with dropbox, you have a folder structure. With Google, it's hidden somewhere.
Totally agree. It makes finding the right document far harder than a product from the company most associated with search in the world should make it.
On the corporate Google Drive version I use at work you can choose to search only within your own drive, but it’s clunky to set up and you have remember to do it before you type in your search term - which is counter-intuitive.
It would be much better if there a selection of filters appeared to apply to the search results - owner, drive, when I last accessed or edited (as you suggested), etc.
(As an aside… don’t folks in Google use their own products and hit the same frustrations? It takes no more than a few days of using corporate Drive to run into this…)
On the corporate Google Drive version I use at work you can choose to search only within your own drive, but it’s clunky to set up and you have remember to do it before you type in your search term - which is counter-intuitive.
It would be much better if there a selection of filters appeared to apply to the search results - owner, drive, when I last accessed or edited (as you suggested), etc.
(As an aside… don’t folks in Google use their own products and hit the same frustrations? It takes no more than a few days of using corporate Drive to run into this…)
Former employee. The internal search, Moma, is actually much worse. Although it was made much better when it was reworked a few years ago. Internal sites can't be ranked in the same way as external ones.
But the truly loved search service within Google is codesearch. And that's great
But the truly loved search service within Google is codesearch. And that's great
Why is internal search in giant companies so hard?
Meta's internal search, bunnylol, is also nearly useless and a big productivity hog.
Meta's internal search, bunnylol, is also nearly useless and a big productivity hog.
Because a big part of the search engine algorithm is recommending things based on how long others spent time there, what other sites they clicked on, etc. There's obv way less data on that for internal search engines.
how to use codesearch outside $googl?
Ah, ok, it only searches Google code, figures... I still have fond memories of the long-defunct Google Code Search and was wondering if they revived it without me noticing, but alas, they didn't...
I read this question differently than other commenters. I think OP is asking if employees can use Google Search to search the internet like the rest of us, but they'd get decent results without the bloat.
It would make sense as it would be in the company's best interest that the employees be shown meaningful results when they're trying to solve problems. A completely different index/ranking algorithm that punishes, instead of rewarding, all the fake SEO crap.
Having said that, I have no idea.
It would make sense as it would be in the company's best interest that the employees be shown meaningful results when they're trying to solve problems. A completely different index/ranking algorithm that punishes, instead of rewarding, all the fake SEO crap.
Having said that, I have no idea.
We get the same Search/YouTube/Whatever when on the Google network or on Google-managed devices. The only difference is that the ads don't generate revenue.
Last I heard everyone at Google uses Bing these days.
Along with iPhones and uBlock Origin.
From the former leaker times, one could see that their censoring mechanism is purely political (CIA, NSA, State Dpt), and less Spam/Quality related. Their censored blacklist is manually maintained, the quality algorithmicly, that's why it can be gamed easily.
We have a parallel copy that indexes our intranet but other than that we use the same one everyone else does.
hey man! I am desperately want to report a series of sites that pure black hat spam the search engine rankings, but its weird that seems nobody (means those famous Googlers on Twitter lol) never responded me regard this, so if possible, if you'd love to see these evil ones and direct these to web spam teams, would be really grateful.
No.
Hey man. If you are from Google, I want to submit you a list of pure black hat spam sites which very deeply disrupted the search engine results page. If you wanna take a look and direct these sites to web spam team, just tell me. (and yes, those Googlers on Twitter ignored me, include John muller)