Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: Is there room for another search engine?

news.ycombinator.com

Post reply on HN
Print view View original

91–100 of 200 posts

Re: Ask HN: Is there room for another search engine?

#91

Post by matt4077 »

dhimes wrote: post #38
Animats wrote: post #24

That's a good question, and something I've spent much time on. Cuil (2008-2010) tried. I knew some of those people. It cost them about $30 million to launch a full scale search engine. They had no revenue model. In retrospect, they were hoping to be acquired by somebody. It was some ex-Google people, trying to replicate older Google technology. They had a great launch, but the system wasn't very good and traffic rapi…

When someone sees a search ad, they're actively looking for the item of interest and may be ready to buy. My experience with Google is the opposite of this. When people are searching google they are looking for information. When people go to Amazon they have their credit card in their hand. Most of my google-ad clicks (85% last check) are clearly not interested in buying what they are looking for. They are only on my…

I'm not sure if it works that way, but if you change to Pay-Per-Conversion, it would be in Google's interest to make ad quality depend on conversions-per-view.

And – as just another anecdote – I have had continuing success over the last 8 to 10 years with search ads, while not having a single conversion with anything else (Facebook, youtube, twitter etc.) as far as I can tell.

Re: Ask HN: Is there room for another search engine?

#92

Post by arekkas »

19eightyfour wrote: post #78
Animats wrote: post #60

Earlier quoted context omitted.

The next direction in search seems to be direct question-answering. Echo (Amazon) is interesting in that, being voice only, it can't punt to a screen of search results. It has to answer the question.

The next direction in search is search that machines can use. Without needing humans to tell them what the results mean.

This isn't task of the search engine, it's the task of the machine that uses it. If the machine doesn't understand meaning there is no way the search engine can teach it that meaning.

So no, this will not be the next direction - unless there is a universal model of machines that understand things. But there isn't, and there won't be for at least another 10 years.

Re: Ask HN: Is there room for another search engine?

#93

Post by jasode »

eridius wrote: post #89
jasode wrote: post #53

If by "search engine", you mean something similar to Google/Bing then probably not. However, if we expand the concept of "search" to something beyond text on webpages and "engine" to something beyond a linear algebra pagerank problem that weighs url links, there's room for many more competitors. Let's say we want to search for "best restaurant": Method #1 might be searching millions of web pages, twitter posts, newsp…

Method #3 would be for calculating popular restaurants, not best restaurants. For example, I bet McDonald's would rate pretty highly with that approach, but it's very nearly as far away as you can get from the idea of a "best restaurant".

Popularity is but one input for "best" -- depending on one's definition of best. There's enough metadata (price, location, cuisine, etc) to correlate with actual payment data to filter out fast food like McDonalds. It doesn't have to be a totally naive statistical approach that gives "dumb" answers.

The idea is that what people actually pay for is a different set of vector inputs compared to what people submit reviews for (Yelp/TA) or what people link to (blog with links to favorite city restaurants.) Google's search index is over 100 petabytes but even that gigantic database is missing lots of data that other entities can collect and convert into uncontested search results.

Re: Ask HN: Is there room for another search engine?

#95

Post by arekkas »

Animats wrote: post #24

That's a good question, and something I've spent much time on. Cuil (2008-2010) tried. I knew some of those people. It cost them about $30 million to launch a full scale search engine. They had no revenue model. In retrospect, they were hoping to be acquired by somebody. It was some ex-Google people, trying to replicate older Google technology. They had a great launch, but the system wasn't very good and traffic rapi…

I was very young when Google became popular. Why did it become so popular? Was it just the technological advance of Google-Matrix (Page Rank) and Map-Reduce?

Re: Ask HN: Is there room for another search engine?

#96

Post by toast0 »

kapauldo wrote: post #32
sergiotapia wrote: post #17

Earlier quoted context omitted.

Algolia is a game-changer. They made it so incredibly simple to add search to your website. I'm not talking about their widgets, I'm talking about their server-side integrations and their javascript client-side lib. It's like magic. See it in action here: https://stackshare.io/match

It's bizarre that algolia is a thing. Elasticsearch is good and free.

I'll give you free, but I'm not sure about good. For the www site I run, we have run a one node instance on each machine; they regularly decide to go off into the weeds and not return results, and in the meantime use ridiculous amounts of ram.

I wish Yahoo would have open sourced some of it's internal search tools. :( They weren't always simple to use, but they were stable.

Re: Ask HN: Is there room for another search engine?

#97

Post by dredmorbius »

sergiotapia wrote: post #17
sairamkunala wrote: post #6

Companies like Algolia which provide a site specific search engine has been doing really well especially with the speed and relevancy where Google currently is not concentrating on. https://www.algolia.com/

Algolia is a game-changer. They made it so incredibly simple to add search to your website. I'm not talking about their widgets, I'm talking about their server-side integrations and their javascript client-side lib. It's like magic. See it in action here: https://stackshare.io/match

It's good in results, but too aggressive in updating live results -- the effect is laggy on systems I use, e.g., when searching HN.

Re: Ask HN: Is there room for another search engine?

#98

Post by dredmorbius »

dri_ft wrote: post #41
stupidcar wrote: post #25

For a general search engine, no, there isn't. The upfront capital investment, in terms of the data center capacity necessary to make a modern scraping and search infrastructure, is immense. And since the ad-word business model does not scale linearly with market share – e.g. the market leader collects a disproportionate share of the available profit – you will be losing additional money for a long time. Since the mar…

> the ad-word business model does not scale linearly with market share – e.g. the market leader collects a disproportionate share of the available profit Why is this so?

Essentially: network effects, efficiencies of scale, and costs of managing multiple small relationships (from the advertising buyer's PoV).

Google's real monopoly at this point is ad-side.

Re: Ask HN: Is there room for another search engine?

#99

Post by dredmorbius »

cocktailpeanuts wrote: post #36

The short answer is YES, but the long answer is, if you're thinking about building a startup, this should NEVER be the question you ask. All successful companies that came out of nowhere and disrupted already stable industry never started out thinking "How do i build another X", "How do I disrupt X"? They all built something they thought was needed by the world and it went onto somehow "disrupt X". So if you're start…

Dunno. It's worked before.

http://infolab.stanford.edu/~backrub/google.html

Re: Ask HN: Is there room for another search engine?

#100

Post by boto3 »

adamnemecek wrote: post #62

I think that definitely. Google lacks in quite a few areas, I think that a.) storing more data about the sites (and doing something interesting with said data) b.) improve the UI/UX for power users. The best part is that I can imagine that there would be quite a few people who would pay actual money for being able to use a better search engine. Note that the Bloomberg terminal, is, among others, a search engine. For…

What kind of code search are you thinking of? In my experience, code search could be useful when one works with a big and unfamiliar code base, but even then good architecture documentation and a good IDE would help more. And when one really needs string search, `git grep` is usually fast enough (for me on a 5GB code base).
Post reply on HN