Overview
Airtable proved that people wanted to build software around their own work. It made a database approachable through familiar tables, then expanded into applications, enterprise sales and AI. Bending Spoons completed its acquisition on September 4, 2026. The product continued operating. [1]
Image 1 / 2
Its trajectory separates product adoption from the financial expectations built around it. The business survived several changes in focus and ownership. The harder lesson is how a broad, popular tool became a narrower enterprise bet, while its valuation expectations fell and its organization repeatedly changed.
Start with familiar actions, then change the model
Howie Liu, Andrew Ofstad and Emmett Nicholas developed Airtable after seeing how often people used spreadsheets to organize things rather than calculate. In Ofstad’s account, the early prototype tested interaction before a full backend existed. Typed fields, linked tables and attachments helped users understand that the rows represented business objects. The team tested privately before broader launches. [2]
Early research began in 2012; Liu’s later transaction announcement dates founding to 2013. By June 2015, Airtable said it had released publicly three months earlier and raised a $7.6 million Series A led by CRV. The ambition was larger than improving spreadsheet formatting: non-programmers could define their own information structures and collaborate on them. [4][13][3]
The trade-off was a long initial build. A generic table alone did not explain the value. Concrete workflows and direct manipulation made the underlying database useful. My inference is that the familiar interface reduced learning cost, while relationships between records supplied a reason to leave an ordinary spreadsheet.
Customers supplied the use cases
In a 2018 AWS interview, Liu said media companies found Airtable for editorial calendars and content production. Word of mouth brought others, and the company built templates after that use case gained traction. This sequence matters: templates followed evidence about how people worked. [4]
Universe made that evidence copyable. At its September 2017 launch, Airtable featured Tomer Sharon’s WeWork UX research system, Jon Gann’s DC Shorts planning, Insomniac’s staffing workflow and Sandwich Video’s production process. Visitors could copy the underlying bases. A customer’s working system became a teaching aid and a distribution asset. [5]
Ofstad also described early pricing as a signal of durability. Customers putting important work into Airtable wanted to know the company would survive. The founder account supports a useful distinction: charging can build trust when people depend on the product, although the price still must fit the value. [2]
Enterprise sales changed the job
Salesforce Ventures’ 2025 account of Liu’s experience distinguishes two selling motions. Early representatives consolidated existing accounts inside companies. Later sellers needed executive sponsors and an argument for making Airtable a system of record for a business function. Organic usage supplied evidence, but it did not automatically create a department-wide contract. [9]
The December 2021 financing funded that expansion. Airtable announced $735 million in new capital, $1.36 billion raised in total and an $11 billion pre-money valuation. XN led the round. The announcement highlighted Interface Designer and its connected-apps strategy. Those figures describe financing and valuation, rather than customer revenue. [6]
In December 2022, Liu announced 254 job cuts and a priority on enterprises with at least 1,000 employees. He said enterprise supplied most revenue and acknowledged that Airtable had tried to pursue too many efforts at once. In September 2023, another 237 roles were cut. That later message described the business as cash-flow positive and said the enterprise focus required a different mix of roles. These were management’s explanations, not audited evidence that the restructuring delivered its intended result. [7][8]
The apparent contradiction between broad adoption and narrow selling is manageable. A flexible product can serve many jobs while its sales team concentrates on a few valuable ones. The operating risk is funding every plausible direction at once. Airtable’s own account supports that explanation more directly than a claim that users rejected the product.
AI changed the creation interface
The company acquired teams as it expanded its AI work. In July 2024, it announced that Dopt would join its AI Pillar, following Airplane and Balsa. Dopt’s own product was scheduled to wind down. The stated purpose was talent and expertise for Airtable’s AI features, rather than maintaining another independent onboarding business. [10]
In June 2025, Liu introduced Omni. Omni assembled Airtable components whose data, logic and interfaces users could inspect. It could also edit records and analyze information. Field Agents extended AI into repeated work across records. The strategic continuity is clear: Airtable still supplied the structured application, while conversation offered another way to build and use it. [11]
Hyperagent was a separate experiment. Liu introduced it as a product from the Airtable team in February 2026. Before the August purchase agreement, Airtable transferred that business line’s assets and liabilities to Hyperagent Inc. Hyperagent’s assets and liabilities were therefore separated from the acquired Airtable business. [24][12]
A continuing business at a lower valuation
The August 2026 agreement valued Airtable at $1.285 billion in enterprise value and approximately $2.25 billion in equity value, including its net cash position and subject to closing adjustments. Enterprise value measures the operating business on a different basis from equity value. Comparing only $1.285 billion with the 2021 financing valuation overstates the like-for-like change. [12]
The gap remains substantial even using equity values. The older financing priced a much larger expected future. It does not follow that every shareholder lost the same proportion: preferences, entry prices, secondary sales and the Hyperagent separation affect individual outcomes, which the reviewed disclosures do not establish.
Bending Spoons reported approximately $480 million in annual recurring revenue as of June 2026, growing more than 20% year over year. Annual recurring revenue is a run-rate measure, not recognized annual revenue or profit. Those disclosed operating figures also challenge a simple collapse narrative. The sale closed in September. [13][1]
The new owner began another organizational change. Product lead Alessandro Savarese’s October 12 letter announced that a large portion of the pre-acquisition team would leave. It promised attention to core features, performance, AI and administrative control, while stating no immediate change to customer plans, prices or service. These commitments are a roadmap; their results remain to be measured. [14]
The present product leaves little room for a clone
As checked on November 5, 2026, Airtable offers a free plan, Team at $20 per editor monthly with annual billing, and Business at $45 on that basis. Read-only collaborators and form submissions do not incur those seat charges. Omni building does not consume AI credits, while many operational AI actions do. Business and Enterprise Scale already include App sandbox for isolated structural changes. [15][16][17]
Baserow offers a free cloud plan and self-hosting, with paid features including role controls and audit logs. Grist also offers a free open-source Community edition. A lower-priced table or a generic AI builder faces established alternatives. [18][19]
A narrower hypothesis follows from the burden of keeping custom apps understandable. An Airtable consultancy could use a portable client-handoff record that combines a comparison of the app’s data structure, manually witnessed user journeys and unresolved checks. Airtable’s partner program confirms that implementation and ongoing support are real services. It does not prove that consultants will buy this particular tool. Native sandbox and On2Air backups already solve adjacent jobs; the proposed product must earn its place through clearer handoffs. [21][17][20]
Timeline
- 2012–2015: Research, private testing and public release established the original product. [4][3]
- September 2017: Universe launched copyable customer workflows. [5]
- December 2021: Series F supplied $735 million. [6]
- December 2022–September 2023: Two job-cut rounds accompanied the enterprise focus. [7][8]
- June 2025: Omni and the AI-native relaunch arrived. [11]
- August–September 2026: Hyperagent was separated, the sale was agreed, and Bending Spoons completed the acquisition. [12][1]
- October 2026: The new product lead announced further team changes and core-product priorities. [14]
Sources
- SEC: completed acquisition
- First Round: Andrew Ofstad on product-market fit
- Airtable: Democratizing the database
- AWS: Howie Liu on customers and costs
- Airtable introduces Universe
- Airtable: Series F
- Airtable: New strategic direction
- Airtable: CEO message
- Salesforce Ventures: Platform evolution
- Airtable: Dopt joins AI team
- Airtable: AI-native relaunch
- SEC: purchase agreement and Hyperagent transfer
- SEC exhibit: transaction announcement
- Airtable: Back to bases
- Airtable pricing
- Airtable AI billing
- Airtable App sandbox
- Baserow pricing
- Grist pricing
- On2Air backup pricing
- Airtable Services Partner Program
- Howie Liu: Introducing Hyperagent