A startup is testing a subscription model for search engines
wired.com
wired.com
Google benefits from the fact that they have billions of users. The nominal cost of their overhead per user is very very low. A company like this is gonna struggle to get even a million users, let alone billions. So even if they charge, it's unlikely they can keep up with Google's infrastructure.
I care about my privacy a little, but less than the cost of a subscription + added hassle / downgraded user experience. I'm betting the the market of people for whom the tradeoff is worthwhile is low enough thast this company won't take off. But I support competition, so I hope i'm wrong.
If I had to bet, Neeva will go after business users and combine Bing's index of the public web with their index of internal business data to make a catered experience that's subscription worthy.
I mean it kind of makes sense. How is there not currently a company offering businesses a way to circumvent Google's ad machine and have customizable scope and indexes?
Completely agree. There is much easier money to be made in a B2B premium search app than in the consumer space.
It is also a common path for search companies; for example FAROO went that direction.
Also maybe they won't proactively tweak results to make political statements. At least in this regard Bing is better, but again they know what you search for.
you forgot the ads!
> it's unlikely they can keep up with Google's infrastructure.
A custom search engine for a specific scope of interests will very quickly get better than google.
Take YaCy, feed just HN into it and have it crawl 2 or 3 pages deep. The level of curation around here isn't modest. Take your own comments and have it crawl a bit deeper for those.
If someone without technical skills can make something like that in a few [low effort] days without any fancy hardware it should be easy to make the proverbial google killer.
Finding a niche worth buying a subscription for is a different story entirely ofc. I imagine it could be bundled with something less ambitious that people are used to pay for that caters to the same niche.
Let's say that's the case (which absolutely isn't true), the challenge is that Google currently controls the access points to the web for most people. Google owns the default settings in Chrome, and they pay Apple and Mozilla billions of dollars a year to also be the default search engine in their browsers/devices.
We hear a lot about SEO spam but the other end of the spectrum seems underestimated: Google likes to dictate how your website looks and behaves. If you fail to live up to those demands you shall either not be indexed or indexing will be limited to a hand full of pages.
It makes a good example in it self. The search results here are dominated by penalized SEO practices. The reality is that you need to stop linking to things google doesn't like.
When I write an article I aim to write quality content. This involves gathering all relevant websites about the topic. I want the reader to have access to ALL pages written by knowledgeable people but also all pages advocating opinions I do not agree with or what I consider misconceptions. Ideally there is a flat list of references and material that creates the full picture of stuff available online.
I have a lot of interests and like to zoom out. My next article might be related to the first in some alien harebrained way only understood by me or it might be about something entirely different.
After just a few of such repeated "offenses" google will already punish the site. One could write 3 of those masterpieces per day for a decade.
You can revisit previous write ups and add even more links. Those will be ones you couldn't find previously! The really "bad" ones. And oh? Look what you've done! You've added to many links to the article.
If you wanted good ranking you should have pooped out one of those shallow turds google loves so much. Short articles, very limited information around a very limited scope and lots of links pointing at your own pages rather than the topic at hand. (the person or website you wrote about)
Articles about SEO penalties of course do not link to the google documentation for webmasters.
"Like Neeva, Startpage sources search results externally—in its case, directly from Google"
I've been having a lot of trouble with Google search results lately. They seem to be getting worse all the time. I'd love to join a new search engine, but if the limit of their search quality is Google, I just can't get that excited.
I'd prefer to start using something that is 80% of the quality of Google and getting better, than something that is fixed to Google.
I can imagine paying a subscription to use it is one way to have a good and unpopular search engine.
Those kind of discrepancies really make it tough for me to get interested in a company, because despite the best intentions of people, function often follows form. If an organization doesn't take their mission seriously enough to modify their corporate structure to fundamentally align the two (I think a public benefit corporation with reporting requirements is an example of that), then I'm not sure why other people should either.
Anybody know anyone over at Neeva? I'd like to talk to them.
DDG and Neeva are both based on Bing, and both integrate (or plan to integrate) many other search sources. Startpage only provides Google results supplemented by Wikipedia for some queries.
The one advantage Neeva might have is that it integrates personal services as search sources (like gmail and dropbox). Personally that has never been a pain point for me though.
A few months back I started working on a search engine (in my sig) based on the hypothesis that a better search engine can be built by curating vertical specific sources to provide higher relevance for niche search topics. I truly believe a federated data source search engine will be the future, but time will tell.
This is true, but only relevant if your goal is to help the most people.
Looking at companies like Superhuman (premium email client for business executives and investors), you can build a company that focusses on power users.
If a search engine were to be optimized for Programers, Product Managers, Team Leads, Directors, etc it could improve lot's of people's work.
DDG is also a second-tier provider that is sourcing its results from Bing.
> We also of course have more traditional links in the search results, which we also source from multiple partners, though most commonly from Bing (and none from Google).
Source: https://help.duckduckgo.com/duckduckgo-help-pages/results/so...
In the end, the company notices that when they make improvements to their algorithms, their revenue goes down. So they stop making improvements.
If it's pay-per-search there is zero friction for me to go to another provider.
It's probably not a practical charging mechanism, but it's an interesting thought. I also don't believe a monthly charging subscription is the right model but this is a tough one to get right.
There would be an optimum point where the results are good enough that people won't leave the service en masse, but bad enough that they spend way more than necessary on it. The optimum point is where the profits will be maximized for the search provider. Thanks to the advances is AI/ML/Data Science methods and tools, a savvy company will find this point and stays there until the point moves.
I first heard this idea from Richard Stallman many years ago.
I think at some point they changed the way it works and it fizzled out. I barely see it any more.
For YouTube: Maybe YouTube Premium is what you need then? No ads and you are supporting the creators you are watching.
Flattr also follows the subscription model. I want the busking model. I provide my payment details but not a penny is spent until I click the button for a specific video/channel/piece. The key is low friction. It can't be some third party thing where you need to click away. It needs to be right there where you choose what to watch/listen to.
anyone know how this works? can any startup just proxy Google like that, or do you need to be some sort of official search partner of Google? just curious bc I have often wondered if I myself could attempt a modified search system as a side project.
Or perhaps a vertical search engine for B2B...
If it is truly a general B2C search engine play, I think ignoring DDG as a datapoint is a misstep.
Try searching for a commercial term. For example: say you want to buy an OLED TV, and want to see what it looks like wall-mounted. So try "OLED65CXPUA wallmount" in Google, and you're barraged with ads and ads-disguised-as-content. Even the Google Image Search page is littered with ads (both overt and covert).
* Aggressively prune bad actors (spammy, zero-value, etc.) with ML.
* I am ok with rated-E only content (no porn, etc).
* Zero ads, I am fine with paying premium subscription price / month.
This dream search engine should have a considerably small index size, thus low operational costs.
For me, DuckDuckHack has always returned really awesome results on top of what I consider to be a fairly lackluster set of results for some searches that bing provides. If another provider wanted to step it up with extra sauce on top, I'd be more than happy to try if it makes my day to day better.
> but I definitely get different results from Google, Startpage, and DuckDuckGo
Different results from Google and Startpage are expected, because Google offers personalized search adjustment. Even if you're not logged in, Google uses Ubercookies and fingerprinting techniques to identify you as a unique user
> > We also of course have more traditional links in the search results, which we also source from multiple partners, though most commonly from Bing (and none from Google).