One company's plan to build a search engine Google can't beat

You can't out-Google Google. For two decades, companies have tried to build a better search engine, and now the internet graveyard is filled with would-be competitors. Even the smash successes only get to 1% or 2% market share. Yet companies keep trying because the opportunity is so large: eMarketer predicted the search-ad market will be $64.72 billion this year, and $86.06 billion by 2023. In that business, even single-digit market share mints a billion-dollar company.

Sign up for Protocol newsletters

Sridhar Ramaswamy knows this better than most. At the end of his 15-year run at Google, he was in charge of the company's entire advertising division. "I joined that team when they were 15 people," he told me recently over Zoom, "and when I left, I was managing a team of over 10,000 people. It was crazy." Given his inside knowledge of the ads business and long experience in search, he'd be a prime candidate to go bite off a chunk of the market for himself.

Instead, Ramaswamy is doing the opposite: He's building a search engine with no ads at all. It's called Neeva, and it's a big bet that some subset of Google's billions of users believe search is important enough, and Google increasingly problematic enough, that they'll pay Neeva's (unannounced) monthly price for a search engine. More than that, it's a big bet that reinventing search has less to do with beating PageRank and more to do with fixing the user experience, starting by putting search results back at the top of search results.

Search is too important to relegate to the bottom of an ad-filled page, Ramaswamy said. "It's one of the few natural consumer aggregation points. That's a fancy way of saying, there are not many places that a very large set of users will reliably go to." And to Ramaswamy, it felt like a product being made worse by its business model. "We saw the pressure that ads put on any business," he said, referring to himself and co-founder Vivek Raghunathan, a former YouTube monetization executive. "We could also foresee a future: Usually there are more ads on a Google search, not less."

Neeva founders Sridhar Ramaswamy (left) and Vivek Raghunathan.Photo: Courtesy of Neeva

He's not wrong. Google has been adding ad slots to its product for years, while also looking for ways to make ads and organic results look more alike. It made one change earlier this year that made the two virtually indistinguishable, but then rolled back the change after everyone from researchers to members of Congress accused Google of misleading searchers. Google is forever stuck between trying to surface the best results and trying to get people to click on ads. Ads usually win.

The Neeva founders figured that by inverting the business model from ads to subscriptions, they could incentivize themselves to build a search engine that prioritized privacy, user experience and simplicity instead of scale, scale and scale. But it's not that Ramaswamy had a change-of-heart epiphany about the ad business. It's just that ads get in the way, and with them gone there's all kinds of room to play in search.

Search is where most shopping starts, for instance, so Neeva has worked on more-helpful ways to present search results rather than renting space to the highest bidder. They prioritize well-regarded review sites, show lots of pictures and let people set price alerts. "The fact that we don't have to show ads," Ramaswamy said, "and we have the entirety of the page as a canvas for what we can imagine? We're super excited about that."

Rather than try to build a search infrastructure from scratch, Neeva instead opted to use Bing's search API for its basic results. They're also working with some of the same sources as Google and Bing, to surface things like stock prices, Wikipedia information, maps and the like. So, fundamentally, Neeva's search results won't be much different than the competition's.

What Neeva's trying to do is innovate on top of those basics. "The thing you get when you work in an area for a long time is the ability to slice and dice the problem, figure out which ones are the most interesting, and go invest in them," he said. He's trying to stay focused on what Neeva's 28-person team can do right now, but he knows exactly how much work it will take to tame the endless realm of search results. Right now, he's focused on shopping in particular, but Neeva's also optimizing Stack Overflow search to create a simpler way for software engineers to get information and plans to approach other niches the same way.

This is where Neeva's plan goes from wild optimism about the future to something almost conniving. See, Google's business model won't allow it to compete with Neeva. It can't get rid of ads because ads are its whole business. Google's locked into a path of seeking ever more growth, finding ways to get users to click on ever more ads, collecting ever more data. Like Hey did with email, Neeva isn't just seeking to do things the tech giants aren't doing. It's looking for things they can't do at all.

Neeva's most unusual feature is its ability to also search users' personal files. In a demo, Ramaswamy searched for tax documents and photos, all surfaced within his search results or available in a Personal tab in the Neeva interface. As people's documents and information get spread around services, Neeva's not the only company trying to bring them together, but again, Neeva's theory is that search is a perfect place to solve this problem. Its bookmarking feature also combines personal and web documents, so users can store everything together. By combining both, Neeva hopes to become the place users go when they're looking for anything, no matter what or where it is. That's how you make your app a habit, how you might actually go about breaking our collective Google muscle memory.

One tricky thing for Neeva to navigate will be how it handles user data. While it doesn't sell ads or user data, it still plans to store information about users to help improve the product. By default, it holds that data for 90 days at a time, but Ramaswamy said he's still thinking through exactly how to handle user data and privacy. With that, too, comes questions about moderation, misinformation, and the role any search engine plays in a user's experience of the internet. Ramaswamy said he doesn't have a grand unified theory of this yet either, but he hopes the business model works in Neeva's favor here, too. "We want there to be a degree of security in you knowing that your data serves you, whether it's your behavioral data, or the things you attach to Neeva."

Ultimately, Ramaswamy said he sees identity and trust as part of the whole Neeva experience. "Take news recommendations," he said. "One worry people have in any signed-in product is hey, are you going into a rabbit hole, will you create an echo chamber?" Maybe Neeva users should be able to share their own news feeds, or pop into others, to experience how someone else sees the internet. But there, too, there's still plenty to figure out

Neeva's in closed beta for now, with plans to launch more broadly early next year. Ramaswamy seems to sense that he's doing this at the right time, as Google comes under more scrutiny than ever before. He figures that he'll be on the right track if Neeva can carve out, say, a million paid subscribers in the next two years — a million people who care enough about the way search works and the way their data is collected and used. He's trying to compete with Google by not competing with Google. Which might be the only way to do it.

from Hacker News https://ift.tt/3jx2YKv