Overview
Neeva attempted to sell a search experience without advertising, built by a team with deep experience at Google. It developed search infrastructure, browser features, and AI answers with citations. It ended consumer search in June 2023, then its team and technology continued at Snowflake. Snowflake later disclosed $185.4 million in cash consideration for Neeva and its equity investee.[1][8][10]
The founders identified acquiring and activating users as a larger obstacle than persuading existing users to pay. That distinction matters: a product can appeal to people who try it and still fail to recruit enough of them. Browser partnerships addressed part of the problem, but did not establish a sustainable consumer business.[4][7]
Image 1 / 1
Founding Story
Sridhar Ramaswamy and Vivek Raghunathan founded Neeva in 2019. They met at Google. Ramaswamy became senior vice president of advertising and commerce; Raghunathan became a vice president in advertising. Their proposition changed the payer: searchers would fund the experience, reducing the need to optimize it around advertising. Sequoia's original investment account also emphasized combining public search with information from private accounts that users chose to connect.[1][3]
In the 2022 founder AMA, Raghunathan wrote, “Search ranking is editorial.” He also described an engineering constraint: “the web is pretty hostile to new crawlbots.” The search engine needed relevance judgments and usable source access alongside its different business model.[5]
The company did not begin without support. Greylock describes early work after Ramaswamy joined as a venture partner in 2018, an initial team housed in its offices, recruitment help, and product-feedback sessions. That earlier ideation is separate from the conventional 2019 founding date. The case therefore tests the limits of experience and access to resources, rather than a lack of technical credentials.[2]
Timeline
- 2019: Neeva was founded, following earlier product exploration.[2][3]
- 2020: Sequoia published its investment thesis for private, user-funded search.[1]
- 2021: A $40 million Series B brought reported funding to $77.5 million at that time. Vivaldi added Neeva as an available US search option.[3][4]
- 2022: The founder's public AMA described free and premium versions, its own crawler, and external search supplementation.[5]
- Early 2023: NeevaAI added sourced answers; Vivaldi promoted broader international availability.[6][8]
- May–June 2023: The founders announced consumer retirement, with service ending June 2. Snowflake announced an acquisition.[7][8][9]
- 2024 onward: Snowflake explicitly connected Neeva's technology to Cortex Search; that enterprise service remains documented today.[11][12]
What They Built
Neeva supplied a traditional search experience with a different funding model, then extended it with browser and AI features. Its infrastructure was more than a reskinned external results feed. In the 2022 AMA, Raghunathan described Neeva's crawler and index, with Bing used where its own coverage was incomplete. He also described ranking work and human relevance evaluation. These were founder accounts, rather than an independent audit of index scale or search quality.[5]
Vivaldi's September 2021 announcement shows a concrete distribution effort. Desktop and Android users in the US could select Neeva in browser settings after creating an account. The offer included a trial followed by $4.95 monthly pricing. By March 2023, the partner described global access, a free tier, and $5.99 monthly Premium with additional benefits. Pricing and packaging evolved; one historical price does not describe the entire business.[4][6]
NeevaAI combined synthesized answers with linked sources. This reduced some reading effort but introduced a verification problem: a visible citation must actually support the attached statement. A 2023 research audit included NeevaAI among four generative search services and found frequent unsupported statements and inaccurate citations across the tested group. The study's aggregate results should not be assigned to Neeva alone.[6][15]
Market Position
Target Customers
Neeva sought people dissatisfied with advertising and tracking in general search. Connected-account search offered an additional personal benefit. The challenge was reaching users before they experienced those differences, while asking them to create an account, understand the offer, and change a routine default.[1][4]
Market Size
Search is widely used, but total internet use is not a serviceable market for a paid challenger. Neeva needed enough reachable users who would try, retain, and pay for the service at a sustainable acquisition cost. The public evidence reviewed here does not provide a complete paid-subscriber cohort, retention curve, or contribution margin. Funding and an acquisition price are different measures from customer demand.
Competition
Kagi currently offers user-funded search: a 100-search free trial, a $5 monthly plan for 300 searches, and a $10 unlimited-search plan, plus sales tax. That continuing offer weakens the claim that paid consumer search is inherently impossible. Its pricing alone does not prove profitability or explain why Neeva's distribution differed.[13]
The enterprise alternative also developed substantially. Cortex Search combines keyword retrieval, semantic retrieval, and reranking. It charges for several resources, including serving an available index even when no query arrives. Its access model uses the service owner's rights, so an implementation needs deliberate permission design. A generic enterprise search rebuild would enter a market with an established continuation of Neeva's work.[12]
Business Model
Subscriptions were intended to align Neeva with searchers rather than advertisers. Free access and trials lowered the initial commitment, while premium features supplied a reason to pay. The model still required expensive search engineering and a repeatable way to bring people into the product.[5][6]
In the closure explanation reproduced by Daring Fireball, the founders wrote that “acquiring users has been really hard.” They said payment was less difficult than getting people to try a new engine. This is management's diagnosis. It does not isolate the effect of defaults from product quality, awareness, competitive responses, or the financing environment.[7]
Traction
The funding rounds, browser integration, operating search service, and AI release demonstrate a serious technical and distribution effort. They do not provide a verified path to consumer profitability. Neeva's engineers also produced assets an enterprise buyer valued, with Snowflake identifying talent and developed technology as its acquisition rationale.[3][4][10]
The evidence supports a narrower conclusion than either total failure or straightforward success. The initial consumer proposition ended; the company and technology had a continuing enterprise use. Original investor and employee returns require capitalization and transaction details beyond the headline consideration.
Outcome and evolution
Snowflake announced its acquisition in May 2023. Its later financial filing records the completed cash stock purchase during the quarter ended July 31, including Neeva's equity investee. The $185.4 million consideration included a business with approximately $44 million of acquired cash. An $83 million developed-technology valuation and goodwill were accounting allocations, rather than additional purchase payments.[9][10]
The enterprise continuation became more concrete in June 2024, when Snowflake stated that Cortex Search used Neeva technology alongside its Arctic embedding models. Current documentation confirms an available search service, rather than merely an acquisition announcement. This changed both the buyer and distribution context: a feature inside an existing enterprise platform differs from persuading consumers to change their daily search habit.[11][12]
Key Lessons
- Test activation as well as willingness to pay. A subscription can appeal to a user who has already tried the service while remaining difficult to distribute.
- A partnership is a channel to measure. Browser integration reduces setup friction, but it does not guarantee that users select and retain the option.
- Validate citations as product behavior. A source link is useful only when it supports the claim a user relies on.
- Separate product retirement from transaction value. Neeva's consumer service ended while its people and technology continued in a different business.
Sources
- Sequoia — Original investor thesis (2020-06-19).
- Greylock — Early ideation and team support (2021-09-13).
- Techcircle — Series B and origin (2021-03-10).
- Vivaldi — US browser partnership and screenshot (2021-09-15).
- Vivek Raghunathan / Reddit — Founder AMA (2022-03-18).
- Vivaldi — Global access and AI search (2023-03-01).
- Neeva / Daring Fireball — Reproduced founder retirement explanation (2023-05-22).
- Search Engine Land — Closure schedule and January AI launch (2023-05-22).
- Snowflake — Acquisition announcement (2023-05-24).
- Snowflake / SEC — FY2026 acquisition accounting (2026).
- Snowflake — Cortex Search technology lineage (2024-06-04).
- Snowflake — Current Cortex Search documentation (checked October 23, 2026).
- Kagi — Current consumer search pricing (checked October 23, 2026).
- Algolia — Current search implementation baseline (checked October 23, 2026).
- Liu, Zhang and Liang — Generative-search verifiability audit (2023).