> They are working out the kinks with the pricing scheme
You mean being way too expensive? I appreciate them wanting to stay afloat but they need to bring the cost per search down drastically. They are doing something wrong behind the scenes with regards to spend or optimization.
> They are doing something wrong behind the scenes with regards to spend or optimization...
I disagree. You have to remember how disadvantaged the online ad market actually is. Many of those ad networks also are simply legitimizing fraud on the advertisers, but because the networks aren't big enough at competitors, and all existing implementations suffer the same issues you really don't get what you pay for as Uber found out years ago.
This is what has been holding me back from trying out Kagi. I will probably jump over to Kagi at some point because I've noticed the quality of DDG searches going downhill as well. However, I'm firmly in Kagi's idea of 1% of Internet users. I search for anything at the drop of a hat. I just checked my history in Firefox for duckduckgo.com for "This Month" and it came back with 704 individual searches (DDG likes to add on a =web and generate a new request so I filtered for just those). My phone is an island all on its own, and I probably have another 100-150 searches on there. March and February are right around the same numbers. For fun'sies lets just say 1000 searches per month. $10/month for the 700 searches plan + 300 * 0.015 for the overages = $14.50/month.
Unlike the other comments down-thread, this number feels right. I don't like it, and my gut doesn't want to accept this. If you're not harvesting user data and/or selling ads, and the monthly subscriptions are the only source of revenue then $14.50/month sounds right.
I'm still riding this fun'sies train so lets just kick some numbers around. I suspect their FAQ might be a little out of date, but lets go with it. https://kagi.com/faq says that, "it costs us about $1 to process 80 searches". 1000 searches / 80 = $12.50, so they make $2 per month from me. I hope that 80 searches for $1 covers all of their OpEx. If it doesn't, then that $2 per month gets a bit more sad. https://blog.kagi.com/status-update-first-three-months states that, "Kagi search is currently serving ~2,600 paid customers." from September 1st last year. Let's say they've grown 10x since then and we're at 26,000 paying customers and they're all me. $52,000 in profit on $377,000 per month in revenue. So they're clearing $624,000 a year on this totally hypothetical scenario.
> I hope that 80 searches for $1 covers all of their OpEx
It does cover only the search and infrastructure cost, not salaries and other operations.
According to my calculations, Google is making $5.60 on those same 80 searches from ads (roughly 7 cents per search), so $1 cost from Kagi is a bargain! :)
Just a datapoint, but I feel like I do a lot of searches… I’m always looking stuff up, and I’m also in the habit of using my search engine as a lazy form of autocomplete for urls. I apparently do about 700 searches a month according to Kagi. If I were on the basic plan, I think that would mean I’d be paying $5/mo. + $0.015/search * (700-200) searches/mo. = $12.50/mo.
For what search is worth to me, that seems reasonable. (Or I guess more to the point, how much I value a relationship where I’m the customer, rather than companies expressly interested in using psychological manipulation to influence my behavior…)
Yeah, I similarly felt like I did a lot of searches every day... and if I had estimated my search usage I would have said easily thousands a month. But I'm also actually in the 600–800 range it turns out.
That said, I think Kagi should consider rethinking their pricing model, because the perception is what matters if it presents as a roadblock.
You are correct, Kagi current users are mostly from the remaining 1%. However that plan was made to attract the remaining 99% Internet users to Kagi (and the plan is more than sufficient for their needs).
I'm skeptical that those 99% would ever be interested in Kagi's value proposition at all. Kagi is a niche product in the contemporary context of ad funded free search engines, and it would have to survive until the other side of the adoption curve before it can reliably sell itself to the average person.
45 USD per month is more than what most people in advanced countries pay for their mobile phone service or broadband connection. Frankly, it's crazy to me. Anything more than 10 USD per month for better search is hard to defend. And yes, I know that Kagi is targeting a _very_ picky audience! B2C is a brutal business.
I easily do dozens of searches per day, occasionally hundreds, especially when debugging obscure issues. It shouldn't be the most expensive subscription I've got by far. Well it could be, but I better not have to go to google or something else to get answers, and it's not quite there yet.
You mean being way too expensive? I appreciate them wanting to stay afloat but they need to bring the cost per search down drastically. They are doing something wrong behind the scenes with regards to spend or optimization.