Hacker News Reader

Comment by voldacar | Hacker News Reader

>So how can we fix the system? The best way to fix the system is to prioritize websites that are there to share knowledge, not websites with their primary priority to make ad revenue.

Different solution:

- Make a search engine that is actually good. Blacklist all SEO spam such as pinterest, allow complex queries, restrict by date, site, regex, etc

- Charge 1 cent per search, or as much as is necessary

Why has nobody done this? What am I missing?

I would totally use this. And if the pricing were correct, you wouldn't need a captcha and could just expose it as an api for bots

View on HN

>Blacklist all SEO spam

That is like saying:

"Electricity problems ? just build a fusion reactor..."

Sure it's the answer but the devil is in the details !

The "easy ones" like pinterest are "easy" but I can guarantee you that finding all the other ones and new ones will be much much harder, especially once you have false positives.

>Why has nobody done this? What am I missing?

The answer is in your proposal:

>- Charge 1 cent per search, or as much as is necessary

90% of people do potentially hundreds of searches everyday. Especially non expert users that practically use the search engine to search for the domain of the sites they need to go too.

The people ready to pay a fixed fee /per search wouldn't be able to cover even a fraction of the infrastructure costs of : "search engine that is actually good"

Then there is the question of what good means.

Google's solution that "outside links" are considered as "votes" that this site is good is what made it good but also what made it vulnerable to SEO.

Do you have any other idea on how do you decide if a result is good or not beyond the literal text search?

>Google's solution that "outside links" are considered as "votes" that this site is good is what made it good but also what made it vulnerable to SEO. Do you have any other idea on how do you decide if a result is good or not beyond the literal text search?

Maybe you could have a "report spam" button next to every search result, and sites that got reported more often would get weighted way down in the results. But whatever google was doing around 2010-2012 is my benchmark for "actually good search engine". Maybe hire some people who worked at google back then and have them write something similar

>90% of people do potentially hundreds of searches everyday. Especially non expert users that practically use the search engine to search for the domain of the sites they need to go too. The people ready to pay a fixed fee /per search wouldn't be able to cover even a fraction of the infrastructure costs of : "search engine that is actually good"

I'm pretty skeptical of this. Ok, so you spend a few dollars per day on search. So what? People spend that much on coffee, and you get way more utility out of an actually good search engine. The market would definitely exist

If there was a “report spam” button, companies would hire people to “report” anything that praises their competitors’ products.

Exactly this ! I used to be head of search for a big eCommerce site, and even with this very "limited-scope-vertical". I was always amazed at the vast differences between people's opinion on "what are good results" it's so personal and ambiguous most of the time.

Even in our small'ish devteam (< 15), there were never 100% consensus on what "query to product-ranking" should be.

Doing this at Google scale I can only imagine is x1000 harder

Step 1 is quite difficult, and step 2 has the downside of disincentivizing people to use it. There are some attempts to make search engines where you pay per month.

Kagi Search.

Reply on news.ycombinator.com