Kagi search engine founder Vladimir Prelovac explains the real costs of ‘free’ search

Prelovac joined Michael “Roo” Fey, Head of User Lifecycle & Growth at 1Password, on the Random But Memorable podcast to talk about various search-related topics. These included the trade-offs that most people make online, such as paying for free services with our personal data and behaviors. The interview also explored an alternate future that bypasses search-based advertising with a combination of private and public search engines (the digital version of a public library!).

For more thought-provoking themes related to search, such as do we care about the quality of information we’re putting into our brains, how to judiciously integrate AI into the search experience, and why ad-funded search might not be as efficient as we think, read the interview highlights below or listen to the full podcast episode.

Editor’s note: This interview has been lightly edited for clarity and brevity. The views and opinions expressed by the interviewee don’t represent the opinions of 1Password.

Michael Fey: Can you give us a little bit of background on what led you to create Kagi? What gaps did you see in the existing search engine market that you wanted to address?

Vladimir Prelovac: One day in 2018, I had a Matrix wake-up moment. I was using Google, as everybody else does in the world, and I was searching for GitLab. I saw a sponsored ad for GitLab at the top and then noticed the organic search result was also for GitLab. Both were the same.

My first thought was: “Poor marketing person at GitLab who had to pay for this nonsense.” And then the Matrix wake-up moment happened when I realized: what about me, the user? I realized my intelligence was being insulted by having to see results that were exactly the same, and that it was a result of the ad-based business model enforced upon me. But I, as a user, had no say in this whole scheme.

I thought, for something as important as search and retrieving information, there should be a user-centric search engine. And strangely, there wasn’t. All the search engines in the world at the time were ad-based.

“For something as important as search and retrieving information, there should be a user-centric search engine."

At the same time, my daughter started elementary school here in the Bay Area. They all get Chromebooks, which are conveniently pre-installed with the entire Google suite, including the search engine. It dawned on me that she will be tracked and profiled from age six to adulthood – that’s a decade worth of her information by the time she’s an adult. That was just not acceptable to me. And again, there was no search engine you could pay for. I had already bought a YouTube Premium subscription for my kids because I didn’t want them to be influenced by ads, but I couldn’t do that for search.

I realized that something needed to be done and that we should have access to information that is in our best interest and not have a third-party intermediary. That’s basically the origin story of Kagi.

MF: I always go to the organic search results even if they’e the same as the ads. It’s fascinating and tiring, honestly, to have to mentally filter out these pieces of information that I don’t want to see.

VP: It’s not just mentally tiring. It’s also time-consuming. There’s a loss of productivity. And a lot of people fall prey to malicious ads that take them to websites that weren’t their intention. The FBI in 2022 posted an official guideline recommending that U.S. citizens install an ad blocker. That’s how bad the situation is. We have an official government agency warning citizens that they should have an ad blocker installed. Think about the consequences if everybody did what the government recommends doing. Google would be out of business.

MF: You mentioned your daughter is going to have this profile built about her. How many people understand how their personal data is being used by popular search engines?

VP: I like to compare it to the notion of healthy food. Thirty years ago, nobody cared about healthy and organic food. Now we all sort of care what we put in our bodies. This concept where we think carefully about how we get our information, and realizing the price that we pay for all these services to be free today – I think will take some time. The change in mindset cannot happen overnight. It took us 30 years to start thinking about what food we put in our bodies. I think it’s going to take at least a decade for us as a society to start thinking about the quality of information we put in our brains.

“It’s going to take at least a decade for us as a society to start thinking about the quality of information we put in our brains."

Today the majority of the web economy is ad-based. All these search engines and browsers are free because they’re directly or indirectly paid by advertisers. And at the same time, those are all trillion-dollar companies. If you stop and think for a second: “How come I’m nominally paying the price of zero, but these are all trillion-dollar companies?” You realize there is something they’re getting that is even more valuable than a monthly subscription. The price you end up paying is full access to your online digital life.

MF: How do you describe that trade-off to people when you talk to them about Kagi and the fact that they are unknowingly making these trade-offs?

VP: In every transaction, there is a currency. Because the price of using a search engine is zero, people assume there is no currency involved because we usually measure value in monetary terms. The truth is that the currency used here is your time. Your mental cycles, your productivity, the fact that every search you make and every website you visit is being tracked.

“The whole purpose of advertising is to change your behavior and make you do things you would otherwise not."

And, last but not least, the behavior change. Let’s not forget that the whole purpose of advertising is to change your behavior and make you do things you would otherwise not. We’re paying a pretty big price when you take all these things into account: lost time, productivity, a profile of every citizen, and the behavior change of every citizen.

There is this wonderful quote by Ian Bremmer, a political scientist. He said: “The fact that we as citizens are getting information from the world’s largest advertising company is my definition of dystopia.” So, it’s not just the individual – society as a whole is paying a pretty big price by outsourcing the information consumption to what is the world’s largest advertising company.

MF: On the flip side of all of that, Kagi is prioritizing user privacy. Can you talk about how it’s doing things differently and what data it’s protecting and how?

VP: Once you change the business model and take advertising and third parties out of the equation, you basically say: “You can pay for Kagi with your wallet, not with your data.” There is no more incentive to collect, misuse, or otherwise do anything with user data. In fact, for Kagi, user data is a liability. We don’t want it. We have nothing useful to do with it. You have to log in for Kagi and create an account because that’s the nature of a subscription-based business model; you somehow have to tie the payment to an account. But other than that, there is really no incentive for Kagi to ever want to touch your data.

“For Kagi, user data is a liability. We don’t want it. We have nothing useful to do with it."

The equation is flipped 180 degrees. With a paid business model, you’re paying $5 or $10 per month for search, and there is nothing between you and the search vendors. There are no third parties, there are no ads, there are no misaligned incentives, there is no conflict of interest. There are a lot of positive side effects for the user once the product becomes user centric.

MF: What are some of those unique features that Kagi provides on top of traditional search engines?

VP: The most popular features in Kagi for our users are the ability to block websites you don’t want to see in your search feed. We call it “search feed personalization”. If you don’t want to see that pesky website ever again in your search results, you just say “block it.” And you can do the inverse. If there are websites you’d like to see more, you can give us a signal you want to see them more. You as the user have the ability to personalize your search experience versus us as a search engine doing that on your behalf, which is how the majority of search engines work.

Another search experience feature that people like a lot is Lenses, where you can create a subset of your favorite site and perform a search only against that index. For example, you have your hobby websites or favorite recipe website; you create a recipe Lens in Kagi with those sites, and you just search through those sites, not the entire web. We have a bunch of preloaded Lenses for academic research, PDFs, programming, whatnot.

I would say those are the two most popular features that are unique to Kagi. We are incentivized to create these kinds of features because people pay for the search experience. They can get search for free elsewhere. So, we have this incentive to constantly innovate on the search experience, not just the search quality.

MF: I’ve attempted to move my family to alternative search engines in the past. And the feedback I tend to get is: “It’s not finding what I wanted to find.” How is Kagi answering that type of feedback?

VP: Kagi is not trying to replace or redo what Google did over the last 25 years. Instead, we’re getting search results from almost every search engine in the world. Plus, we have our own index that focuses on what we call the small web, which are personal blogs and websites that are kind of lost in mainstream search engines. We combine all that information together and produce the results for the user.

On top of that, we do our own re-ranking. For example, we down-rank sites that have a lot of ads and trackers because that correlates with the quality of content, and we promote those that don’t. By combining all those sources, we actually have better search results than any of these search engines independently. If you cannot find something on Kagi, you cannot find it anywhere. This aggregation plus our own search index makes it uniquely high quality.

MF: How do advertisements affect the quality of the search results?

VP: There are many layers of why ads do not belong in the context of information retrieval. First you have malicious intent, and you have society’s most vulnerable, least tech-savvy people falling prey to these sorts of scams. The second layer is, it’s mentally tiring. Most of the time, ads don’t do anything useful for you as a user, so you have to either skip them or install an ad blocker.

But there is more nuance to it. The ad-based search engines are incentivized to keep you on them as much as possible. In other words, you may spend time searching for something over and over. Try a search, try a website, go back because you didn’t find what you need, try another one, go back, change your search. You’re spending a lot of time interacting with the search engine, which is what a search engine wants because every interaction is an opportunity to show you more ads.

“I want you to spend the least amount of time possible using Kagi."

In Kagi, the metric I care most about is time to information. I want you to spend the least amount of time possible using Kagi. The first hit in most searches should be the one you need. If you are not achieving that, you’re not going to be paying us. We have a very strong incentive to give you what you want. And our users report, when they go through their search history coming from, say, Google to Kagi, the number of searches they make in a month almost halves. You become 50% more efficient because now you don’t have to do those repeat searches. It’s a pretty measurable and direct impact.

MF: Let’s play buzzword bingo for a minute. AI is playing a larger role in search engines today. How do you see AI changing the search landscape, and how is Kagi integrating AI to improve search results while still maintaining user privacy?

VP: That’s a great question. I’ve grown up with this vision of AI in science fiction movies – you talk to a computer and all that, and I think that’s a pretty cool vision.

Unfortunately, the current generation of LLMs (large language models) is not on track to achieve that. They’re vulnerable and limited in many ways. They’re pretty powerful and they’re able to do a lot of mundane tasks. Rewriting and summarizing content is one of them, and I think this is where they are potentially most useful in the context of information retrieval, which is where search operates. We have been implementing some of these things very carefully and trying to be very mindful because we understand the limitations of these tools. We see AI as a tool, as a hammer, not this cure for cancer.

Kagi has a lot of AI functionality, but it’s all opt-in, it’s all on-demand. For example, if you make a query in Kagi end with a question mark, that will trigger an AI summary of the search results. Otherwise, there is none. It’s just a search engine. We also have the assistant, which is basically a front-end to all the LLMs in the world, where the results are grounded in Kagi search. You get the best of both worlds, access to all these LLMs, but also the information is grounded in our search results.

So, while we are experimenting with this technology like everybody else, we are also being mindful of its impact, its limitations. We are trying to be very thoughtful about how we integrate it.

MF: What does the future of search look like to you, especially if we enter a world where people are becoming more privacy-conscious and aware of the ad-driven search results?

VP: I think what will happen to society is the same thing that happened with the healthy food example I gave in the beginning: that we will grow more aware about the value of information, both as individuals and as a society.

I think a couple of things may happen. One is that in five to 10 years, all sources of information that are ad-based will have a label for entertainment purposes only. That is really the value. There will be paid search engines like Kagi where you will be able to pay to have better search results that will make you more productive and more competitive in the world.

“There should be a public search engine."

I also think that there should be a public search engine. Imagine search.gov. Because access to information is so important, it should be non-discriminatory. Not everyone can pay for a paid search engine, but everybody should have access to information in their best interest. That’s where I see the role of a public search engine. And by the way, we had that for centuries; it’s called public libraries. Unfortunately, they became obsolete about two decades ago and we did nothing to replace them.

In five to 10 years, I hope that we as a society grow out of dependence on the world’s largest advertising company for information and move towards more user-centric search.

MF: For folks who are concerned about privacy while browsing, what practical advice can you offer?

VP: Kagi is also a creator of a browser called Orion. It’s available on macOS and iOS. It is completely privacy-respecting and is also zero-telemetry, which is pretty unique for a browser these days. Zero telemetry means that it doesn’t send any user information to the browser vendor. If you are on macOS or iOS, try Orion. Orion has a free version, or you can pay to own your browser. There are no third parties involved in making Orion.

Other than that, use an ad blocker. uBlock Origin is a great one. If your browser supports it, that’s going to get you pretty far in protecting you from the hostile web, but not against the browser itself. For that you’ll have to have a zero-telemetry browser, so do your research and see what’s available on your platform.

It’s not just about searching the web, it’s about consuming the information from the web, and the search engine is just one part of that. We are also working on a web browser to create this ecosystem of tools that are user-centric. That’s pretty important when you are talking about search and browser, two of the most intimate tools we use every day. Most people don’t realize that their browsing and searching is paid for by a third party at this time.

MF: Where can folks go to learn more about you and Kagi? How do they free themselves from the shackles of ad-driven search?

VP: The easiest is to visit kagi.com. You’ll find information about all products there, the search engine, the browser, and other tools, like the small web initiative that I mentioned. That’s a great starting point to get yourself educated about this whole new world.

Jenn Marshall

Contributing Writer