It’s not easy to take on Google’s search engine. Just ask the startups that have tried.
(Bloomberg) — It’s not easy to take on Google’s search engine. Just ask the startups that have tried.
As the US Department of Justice works to convince a judge that Alphabet Inc.’s Google has built an illegal monopoly in the internet search market, prosecutors are centering their case around the stories of upstart firms that have struggled in vain to carve out a presence in the market. One such company is Neeva Inc., which was founded by former Google executives and launched to great fanfare in 2019 only to abruptly shutter its product earlier this year.
Taking the stand Monday and Tuesday in federal court in Washington, Neeva co-founder Sridhar Ramaswamy recalled how the company thought it could deliver a better search experience by charging consumers for a subscription, rather than serving up ads, which he believed had gradually corroded the quality of Google’s product. But Ramaswamy and his co-founders failed to gain traction with users, and finally decided that they couldn’t build a viable business, even after infusing the product with new AI tools that could deliver short responses to questions.
“People who tried our AI experience genuinely loved it. It was a better, easier, sleeker experience,” Ramaswamy said on the stand Monday. But an economic slowdown, coupled with Google’s paid-for placement on smartphones, meant Neeva couldn’t “grow our subscriber base fast enough,” he added.
Read More: Microsoft CEO Says Google Search Dominance to Give It AI Edge
That hasn’t stopped some search startups from continuing to try, especially as new technologies create opportunities to upend the way consumers use the internet. The rise of OpenAI’s ChatGPT has sparked talk in Silicon Valley about a new model of search in which artificial intelligence-based software will deliver decisive answers to users’ queries, rather than the familiar page of links that Google popularized. Google has already mobilized teams internally to reinvent its search engine for this new era, but some entrepreneurs argue their startups are better positioned to catch the wave.
“We are moving towards a new segment, a new kind of internet where you are basically getting answers served to you instead of links,” said Aravind Srinivas, a former Google researcher whose startup, Perplexity AI, offers a conversational web-search product. “And that’s a market that will not be dominated by Google,” he said in an interview.
Even for those offering next-big-thing search experiences, going up against Google will be a daunting task. On the stand, Ramaswamy said one of the biggest challenges that Neeva faced was Google’s status as the default search engine on the most popular browsers. The company pays billions of dollars to secure spots on Firefox and Apple Inc.’s Safari, an arrangement that is at the heart of the government’s monopoly allegations.
Neeva sought to be added to a shortlist of search engines that users can choose in Safari, Ramaswamy said, but the conversations weren’t fruitful. The startup also had some promising talks with wireless carriers, which help decide what software comes preinstalled on Android phones, but “the complex nature of contracts that they had with Google made it very, very difficult for them to offer Neeva as even an option,” Ramaswamy testified.
The judge asked Ramaswamy why Apple didn’t add Neeva to the choices for Safari. “I honestly don’t know,” he said. “There is not a clear process.” Ramaswamy said Neeva was willing to pay Apple for an agreement if necessary. “It’s one of these situations where you spend sometimes months negotiating and things peter out. I’m well connected with Apple but that doesn’t mean I can get answers.”
His company ultimately decided that the best way to put its technology in front of consumers was via a mobile app, which required considerable engineering resources. Eight of Neeva’s 50 employees focused on building and maintaining the app, Ramaswamy said.
Neeva’s search engine incorporated machine learning and AI tools that the company licensed for several million dollars, Ramaswamy said in testimony Tuesday. In January, the company launched Neeva AI, a search tool that could generate short written answers to search queries — like ChatGPT but less likely to be “making up stuff,” he said.
Neeva AI demonstrated “that there’s a better way to search and a better way for users to get the answers they were searching for,” he said.
The company planned to be self-sustaining with about 20 million subscribers and $200 million to $300 million in revenue, Ramaswamy said. Yet several years after the company’s founding, Neeva was bringing in less than $1 million in subscription revenue, and facing a challenging climate to raise additional capital from investors, Ramaswamy told the court. Seeing little path forward as a consumer app, the company sold itself in May to corporate software maker Snowflake Inc.
“The problems we would face were predictable four years ago,” Ramaswamy said. “Acquiring users, getting them into a habit is something that is tricky. Every startup founder has dealt with that. There is always a race to create a better product fast enough before your employees tire or your money runs out.”
Read More: Google Will Expand Generative AI Tool to Select Sites on Web
With Neeva out of the race, the search startups left standing are still feeling the sting of competing with Google. AI researcher Richard Socher said he founded You.com to offer users a new form of search that would take advantage of the rapid advances in natural language processing, a branch of artificial intelligence that helps machines understand language.
Socher said in an interview he often hears from users that the quality of Google search has deteriorated, plagued by sites that game the company’s search algorithms to jump to the top of the results.
“Some users are conditioned to go to Google and scroll down right away,” he said.
Yet when these users make their way to You.com, they don’t always stay. Many users who try to make You.com their default search engine get lost in the maze of settings necessary to do so, Socher said.
“You have to be tech-savvy to have a choice in the search engines, and that I think is partially by design,” Socher said. To have a better shot at reaching users, You.com had to build a mobile web browser in addition to its core search product, he said.
Still, Socher is starting to see glimmers of hope. You.com introduced what it calls a chat-first search engine, a product that lets users ask AI software questions that help it home in on exactly what they are looking for. Since the product’s launch, You.com’s pool of users has expanded considerably. The company saw a 35% jump in September from a month earlier, Socher said.
Younger users are helping to drive the shift. Srinivas, of Perplexity, said he recently spoke with a tech executive who had learned about the product from his son.
“The lock-in effects Google has are much lower for the younger generation,” Srinivas said.
Since its launch in December, Perplexity has released new features at a brisk clip. Users can now upload documents to have them summarized and also group key queries into “collections” that they can revisit and share with others.
Perplexity’s traffic remains a fraction of Google’s, but Srinivas said he sees himself as competing not with the main search engine — but with the Bard chatbot, the technology giant’s answer to ChatGPT.
In light of Google’s massive user base, Srinivas predicts that the search engine isn’t likely to fade away, but will gradually become less relevant, following a similar arc as Yahoo’s online portal in the early 2000s.
“That’s why the most important thing to do is to build the next big thing,” Srinivas said.
(Updates with additional details from Ramaswamy testimony beginning in 10th paragraph.)
More stories like this are available on bloomberg.com
©2023 Bloomberg L.P.