Hacker News Reader

Comment by freediver | Hacker News Reader

There are plenty - Kagi, Brave, Mojeek, Yandex, Rightdao, Gigablast...

Been using Kagi for the last week and the search results have been SURPRISINGLY good. Like, in the "I don't have to scroll to find it" category of good, and no having to deal with spam sites that just copy the actual answer but somehow rank higher than the original, like currently plagues Google.

This is probably because Kagi uses Google and Bing indexes.

That revenue model is crazy, no way will they last.

> That revenue model is crazy, no way will they last.

Why? Their plan is to be an amazing SE, but only for the comparatively small amount of people that are willing to pay. They aren’t VC funded but boostrapped, so they don’t need to ruin a good thing for dumb returns.

I believe that their pricing will keep their potential customer base too small to remain commercially viable. I'm sure you've seen a supply and demand curve before. The optimal price point isn't where the price is highest, because then no one buys the product/service. It's actually where sufficient customers can afford the product/service to maximise profit. There are very few people willing to pay $30/month for a search engine. Even if Kagi can convince 1,000 people to do so, I do not believe $30,000 a month in revenue will even keep the lights on.

It is highly likely that $5/month, for example, would attract millions of users, and $25 million/month would certainly keep the lights on.

$30 is not the price unless you need a lot of searches. Looking at my consumption tab, I’m usually below $10/month, so I’ll probably end up in a $10-$20 plan. Some napkin math gives around 2,550 searches per month at $30, Vlad said in Discord $30 is for people who want around 100 searches every day.

Also, you can’t forget that every search actually costs them money, as every single search incurs the fee from Google and Microsoft for their APIs. I think there was an idea around having a $5 trial plan with only 10 searches a day or something. Unlimited searches for $5 would be "We are losing money on every customer, but we are making it up in volume"

>Also, you can’t forget that every search actually costs them money, as every single search incurs the fee from Google and Microsoft for their APIs. I think there was an idea around having a $5 trial plan with only 10 searches a day or something. Unlimited searches for $5 would be "We are losing money on every customer, but we are making it up in volume"

Are they really querying Google and Microsoft for every search? This seems highly inefficient. Then Kagi is basically just a fancy wrapper. I thought their ambition was to build out Teclis and TinyGem to serve more and more content and rely less and less on other search providers. At the very least, I don't think they need to hit Google every time someone queries "games." This is hopefully queried once per period and stored in Teclis/cached.

To be blunt, if this is just a fancy wrapper/aggregator then there is no reason to use this over you.com.

It uses data from both Google and Bing, but reorders it substantially. In addition, they use data from their own index and some different sources. The interesting part is the ranking.

And I mean, it should be easy to compare. For me, kagi is substantially better than any other SE I tried, so I’ll pay for it when it releases, but just test it yourself.

We are not optimizing for revenue but for staying independent and sustainable. (Kagi dev here)

In other words it does not matter if we could have $100,000/month with another price point if it would cost us $150,000 to do so (every search has a fixed cost that does not go down with scale).

Our current price point includes a tiny margin that would allows us to break even at around 50,000 users. That is what we are optimizing for.

If I may ask, why do you believe that having more users and profit would undermine your ability to stay independent and sustainable? I would have thought that more users and profit would do the opposite: secure Kagi's ability to stay independent and sustainable. Intentionally reducing users and profit doesn't make any sense to me, given your stated goal.

I think you were arguing for optimal price point which would maximize revenue, not profit?

Our profit margin is already razor thin, I do not think it can be further optimized. Meaning we can not further reduce price to get more users without operating at a loss.

I see the confusion. Optimal price point optimises for profit: https://www.accountingtools.com/articles/2017/5/13/optimal-p...

It sounds like you have a high marginal cost. As in, you pay every time someone searches for something. Can this not be reduced over time with greater efficiency of scale?

FYI I use Kagi and like it. I’ve replaced Google :)

Nope, the cost is pretty much fixed. What we charge the users is just a hair above it so we have a hope of reaching sustainability at around 50,000 users. In other words there is nothing to optimize further.

And it is only about one cent to search the entire web in 300ms with everything else Kagi does, that has to be impressive ? :)

Very impressive :) So if Kagi profits $x per search, wouldn't they profit more with more searches? More users, more searches, more profit, more sustainable.

Yes, but we can not get more users by reducing price as it would not be $X any more, but 0 or -$X per search.

Gotcha! Thanks for explaining, and thanks for the cool work you guys are doing.

And even if you disagree with their ranking, you can change ranking weights for domains yourself.

I guess we need an up-to-date wikipedia page about all those alternatives.

Kagi uses both Google and Bing index.

Doesn't Brave also use Bing underneath?

If their about page isn't lying, about 90% of their results are original.

They bought the cliqz team and their work / tech when that died, so I'd assume they're using what those people had been building.

Reply on news.ycombinator.com