.png)
.jpg)
Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Airtable: The SaaSPocalypse And A Hail Mary
By Ajit Ghuman & Akhil Gupta · August 2026
The Airtable sale, explained
Airtable grew 20% a year and served 80% of the Fortune 100.
It sold for less than it raised.
The SaaSPocalypse now has a clearing price.
The Deal
WHAT ACTUALLY HAPPENED
On August 4, 2026, a software company used by 80% of the Fortune 100, with roughly $480 million in annual recurring revenue growing more than 20% a year, agreed to sell itself for $1.285 billion in cash.
Airtable raised more than $1.4 billion from investors over its life. At its peak in December 2021, it was valued at $11.7 billion. The sale price is roughly one tenth of that peak. The buyer was Bending Spoons, the Milan-based operator best known for buying Evernote and then sharply raising its prices. No strategic buyer like Salesforce or Microsoft came in with a higher offer, and neither did any growth-stage private equity firm.
None of the usual explanations apply. Airtable was not shrinking. It was not losing money: founder and CEO Howie Liu said as recently as January that the company was generating cash, with roughly half its raised capital still in the bank. It was not a niche tool. And it still sold for roughly 2.7 times its annual recurring revenue.
If you operate a SaaS business, this deal is worth studying. We wrote a book about the collision between AI agents and seat-based software, Monetizing Agentic AI, and the Airtable sale plays out its two central ideas in a single document. The first is the SaaSPocalypse, the market-wide repricing of seat-based software that hit this year. The second is the Hail Mary, the bet-the-company rebuild that incumbents like Intercom and HubSpot are running to survive it. Airtable shows what happens to a company that starts the rebuild and stops short of finishing it.
Bending Spoons is paying $1.285 billion in cash for Airtable's operating business, about 2.7 times its $480 million in annual recurring revenue, or ARR. The deal is expected to close later this year.
The $11.7 billion peak came from a December 2021 Series F round led by Thrive Capital and Coatue. Earlier this year, Airtable shares reportedly traded on secondary markets, where investors buy and sell stakes in private companies, at around a $4 billion valuation. Bending Spoons is paying about a third of even that.
The deal also excludes Airtable's agent business. According to the SEC filing, Airtable transferred the assets and liabilities of its "Hyperagent" business line into a separate company, Hyperagent Inc., before signing. Hyperagent is the agentic AI product Liu had been building in public for the past year. Bending Spoons is buying the core SaaS platform. Liu keeps the agent company.
The low price and the carve-out come from the same source. Both are responses to what happened to software valuations over the past twelve months.
The Market
THE SAASPOCALYPSE
Between mid-January and mid-February of this year, roughly $1 trillion in market capitalization drained out of software stocks. Over the trailing twelve months, the figure was closer to $2 trillion. The S&P North American Software Index posted its worst monthly decline since the 2008 financial crisis. The iShares Expanded Tech-Software ETF fell more than 30% from its September 2025 peak.
For the first time in the modern era, software companies also traded at a discount to the S&P 500. For two decades, investors had paid a premium for software stocks because recurring revenue and high margins made them safer than the average business. That premium disappeared. The market was signaling that it no longer trusted the business model.
A Jefferies trader named the crash the SaaSPocalypse, and the name stuck. The root cause was AI making software too cheap to build. Products that once took hundreds of engineers and years of work can now be rebuilt in months by small teams, and cheap creation removes the scarcity that software valuations rested on.
The same efficiency reaches the customer's side. When agents do the work, companies run with fewer people. If an AI agent can manage pipeline, the customer needs fewer Salesforce seats. If it can resolve support tickets, fewer Zendesk seats. If it can run sprints, fewer Jira seats.
Per-seat licensing, the pricing model most of the $300 billion SaaS industry runs on, rests on one assumption: revenue grows with customer headcount. AI-driven efficiency pushes headcount down, and seat revenue falls with it. Even Orlando Bravo, co-founder of Thoma Bravo, the largest private equity buyer of software companies, told his investors in March that the valuation declines were warranted.
A correction reprices growth expectations. The SaaSPocalypse repriced the business model itself: the specific kind of software company that sells seats to humans, charges for access instead of results, and assumes headcount growth is the primary revenue driver.
Destruction is only half of the story, and this is where the image at the center of our book comes in. We describe this moment as an Ouroboros, the serpent that eats its own tail in order to renew itself, because the same force that is killing seat-based software is creating a new financial species.
Agentic companies sell completed work, and that changes which budget pays them. Traditional software is paid for out of the software budget, which is usually about 2% of a company's revenue. Completed work is paid for out of the labor budget, which runs 20 to 40% of revenue. The pool of money available to agentic companies is roughly ten times larger.
Investors already price the two kinds of companies far apart. AI-native platforms fetch 25 to 30 times revenue in funding rounds and acquisitions, while private SaaS deals close at 4 to 5 times. Airtable just sold at 2.7 times ARR, below even the SaaS band, while still growing more than 20% a year. The repricing that started in public stocks has now reached private acquisition prices.
The Buyer
WHO BUYS SOFTWARE DURING A SAASPOCALYPSE
The price makes more sense once you understand the buyer.
Bending Spoons behaves like a private equity fund with one crucial difference: it never sells what it buys. The company listed on Nasdaq on July 1 at an $18.4 billion valuation, raised $1.68 billion, and told investors it has identified roughly 1,000 acquisition targets. Its portfolio already includes Evernote, WeTransfer, Meetup, Vimeo, Brightcove, Eventbrite, and AOL - over 500 million monthly active users and more than 9 million paying customers.
The playbook is public, documented, and remarkably consistent. Bending Spoons buys a well-known brand with stalled growth at a steep discount, rewrites the technology, centralizes the infrastructure, and cuts most of the staff. WeTransfer lost roughly 75% of its team within weeks of closing. Of the 1,830 employees absorbed in the AOL, Eventbrite, and Vimeo deals, the company expects only a few hundred to remain by the end of this year.
Then it raises prices on the existing customer base. Evernote's personal annual plan jumped about 86% after the takeover, with Forbes reporting some plans going from roughly $100 to $249 a year. The free tier was capped at 50 notes in total for the life of the account.
The results say the strategy works. Evernote's revenue grew 34% in 2024 and another 30% in 2025, even as its subscriber count shrank. The growth came entirely from higher revenue per remaining subscriber. Bending Spoons generates $2.57 million of revenue per employee, a figure that beats Apple. And nearly half of its subscription revenue comes from customers who have been with its products for five years or more.
That last statistic reveals the actual asset being acquired: switching costs. Bending Spoons buys products that are painful for customers to leave, because years of data, workflows, and habits live inside them. The company assumes the business will slowly shrink, and it sets the purchase price so that cash flow from the loyal, locked-in customers pays for the deal several times over, even as part of the base leaves in protest.
Private equity had already become the dominant consolidation force in SaaS; Q1 2025 set a record of 73 PE-led enterprise software transactions. Bending Spoons is the industrialized, permanent-capital version of the same trade. We call this the harvest phase of the SaaSPocalypse. When a business model is structurally impaired, the winning buyer is the one who does not need the model to keep working. Bending Spoons only needs the invoices to keep going out while prices go up.
That is who bought Airtable. So why did Howie Liu, a founder with cash in the bank, a growing business, and no forced timeline, sell to them?
The Play
THE HAIL MARY
Liu's other option was the move we spend the second half of our book teaching companies to run. We call it the Hail Mary: a bet-the-company rebuild for the moment when AI agents can do the work your customers currently do inside your software.
The play runs in a strict sequence. The company rebuilds its product around an agent that performs the work. It reprices around the work performed instead of the seats occupied. It rewires billing, sales compensation, and finance around consumption and outcomes. It retrains the sales motion around the new metric. Then it migrates the existing customer base onto the new model, which is the riskiest phase.
Our core argument is that this play, with your existing revenue as the wager, has become the default. The seemingly prudent path of waiting, piloting, and forming committees is the one that kills you. The math is plain: when your AI makes customers need fewer seats, your product's success shrinks your own revenue.
Intercom ran the violent version of the play. Eoghan McCabe returned as CEO in late 2022, cut deep, and pointed the entire company at Fin, an AI agent that resolves customer support conversations on its own instead of helping human agents resolve them. He then rebuilt the entire pricing architecture around it. Fin charges $0.99 per resolution - the customer pays when the issue is actually solved, and does not pay when it is not. Seats became a platform access fee; outcomes became the growth engine.
Intercom then reworked its legacy pricing and knowingly gave up tens of millions in existing revenue to force the migration. McCabe wrote that the only way forward runs through the destruction of your past. By May 2026, the transformation was so complete that the company renamed itself Fin.
HubSpot ran the staged version. Yamini Rangan moved a public company doing $3 billion in revenue through three visible phases over two and a half years. First, in 2024, HubSpot bundled AI into every tier under the Breeze brand. Next, in 2025, it introduced a consumption currency called HubSpot Credits. Then, in April 2026, it made the leap to outcome pricing: $0.50 for every conversation its Customer Agent resolves and $1 for every qualified lead its Prospecting Agent finds. Each price attaches to a finished result.
HubSpot's $0.50 resolution also undercuts Fin's $0.99 by half. A price war for AI-delivered outcomes has already begun, and seat-sellers are not even at the table.
The two companies have different temperaments but the same destination, and both kept their customer bases while transforming the revenue underneath them. That detail matters for Airtable, because Airtable started down the same road.
THE BOOK BEHIND THIS BRIEFING
MONETIZING AGENTIC AI
The SaaSPocalypse, the Hail Mary,
and the full playbook behind them
Ajit Ghuman & Akhil Gupta
First Edition, 2026 · Monetizely
The Missing Step
THE HAIL MARY AIRTABLE LEFT UNFINISHED
Howie Liu saw the shift coming. He had already cut nearly 500 roles across two rounds in 2022 and 2023, and he described his own posture over the last two years as wartime leadership. He then ran an aggressive AI pivot, a restructuring he described as refounding Airtable as an AI-native company.
The timeline shows how hard he pushed. In June 2025, Airtable launched Omni, a conversational agent that builds Airtable apps, alongside a full repositioning as the AI-native app platform. In the fall of 2025, it acquired DeepSky, an agent startup, and hired David Azose, who led engineering for ChatGPT's business products at OpenAI, as chief technology officer. In January 2026, it launched Superagent, the company's first standalone product in 13 years, a multi-agent research system that Liu summed up in one line: "You're not prompting an AI. You're orchestrating a team." TechCrunch reported at the same launch that Liu viewed the collapse in Airtable's valuation as just the warm-up, with the company's biggest moves still ahead.
Weeks later, Superagent gave way to Hyperagent, an even bigger swing: cloud agents with their own computing environments, deployable into Slack as intelligent coworkers. Six months later, he sold the platform.
Measured against the Hail Mary sequence, Airtable ran the product step at full speed and never started the pricing step. The core platform that 500,000 organizations run on kept its per-seat plans. The agent products launched beside the legacy business - separate brands, separate domains, separate teams - and even they were priced conservatively. Superagent debuted at $20 to $200 per user per month with bundled inference credits, and Liu was candid that the company was not optimizing for margin.
That pricing cut against two rules we argue for throughout the book. Credits are only a transition step on the way to value-based pricing, because customers cannot feel what 100 credits are worth the way they can feel $0.50 per resolved conversation. And per-user pricing on an agent product quietly brings back the seat-based anchor the agent exists to break. HubSpot treated credits as a bridge and moved to outcome pricing within a year. Airtable stayed on credits.
The skipped step also left the two product lines working against each other, in exactly the way the Ouroboros describes. Omni builds Airtable apps through conversation, which is work that today happens in paid seats. Hyperagent deploys agents into Slack as coworkers that complete tasks people currently handle themselves. Every advance on the agent side shrinks the seat count the core platform bills, and the core was never repriced to capture what the agents create.
Intercom and HubSpot broke that loop by changing what the core charges for. Airtable broke it by cutting the company in two. The legacy platform - the seats, the workflows, the renewal base - went to the one buyer whose entire model is monetizing exactly that. Liu kept the clean-sheet story: Hyperagent was already courting the next generation with $20,000 in inference credits for each of 500 agent-first founders, a $10 million commitment announced months before the sale.
Intercom transformed its base. HubSpot migrated its base. Airtable sold its base and kept the bet. The warning we make in the book held to the end: a company that does not reprice its own core eventually gets repriced by someone else.
Was this a failure? In one sense, no. Liu converted a structurally impaired asset into $1.285 billion in cash. He also kept an agentic company with no legacy P&L to defend, no seat revenue to cannibalize, and no migration war to fight. He marked the company down voluntarily, from the inside, at a price he chose before a distressed process could choose one for him.
In another sense, yes. The transformation path, where it works, is worth more. Intercom kept thirteen years of customer relationships and converted them into outcome revenue: Fin grew from $1 million to more than $100 million in revenue and now resolves over a million issues a week. Airtable monetized thirteen years of customer relationships at roughly a tenth of the 2021 price, and now starts nearly from zero in one of the most crowded categories in technology, competing with the very companies Liu himself has named as the only others with real agent architectures.
The carve-out is a rational trade. It is also a smaller one.
What's Next
WHAT HAPPENS TO THE PLATFORM NOW
Bending Spoons' announcement promises improved monetization and more AI. AI plays a different role in that plan than in a Hail Mary. Intercom and HubSpot use AI to change what the customer pays for. Bending Spoons uses AI to run the product with less headcount and to justify higher prices. If you are an Airtable customer, budget for your renewal accordingly.
For Operators
THE QUESTIONS THIS DEAL LEAVES OPEN
The deal asks questions that only you can answer for your own business.
What does "safe" mean now, if a category creator growing 20% a year sold at a harvest price? Would a buyer looking at your company believe your revenue survives agents?
Which ending are you heading toward: transforming like Intercom, migrating in stages like HubSpot, or selling into the harvest like Airtable? Is waiting a real fourth option, or the third ending at a worse price?
How far has autonomy come in your category? If agents are moving from assisting the human to doing the work, how long does your pricing metric have left?
What is your company worth as a harvested annuity, run for cash while it slowly shrinks, at 2 to 3 times ARR? What would it be worth repriced around outcomes? Who captures the gap if you wait?
Airtable's board faced these questions and took the certain number. Your answers are still open.
The Close
THE SNAKE, AGAIN
The Ouroboros image from our book fits this deal because Airtable held both ends of the loop. The sale cut the loop in two. The platform being consumed went to a buyer who will run it for cash. The agents doing the consuming stayed with the founder.
Every piece of this deal is a trend we documented in the book: the repricing of seat-based software, the rise of the harvest buyer, the cost of an unfinished transformation, and the founder betting on agents. The sale collected them in one transaction.
Software will be fine, just as the internet was fine after the dot-com crash. What is ending is one specific kind of software company. The companies that make it through will define the next twenty years, and every one of them will have made the move before the market made it for them.
The question for every operator is which side of the purchase agreement you plan to be on.
FROM THE AUTHORS
Do not run the Hail Mary alone.
Ajit Ghuman and Akhil Gupta are the co-founders of Monetizely, a pricing strategy consultancy for B2B SaaS and AI companies, and the authors of Monetizing Agentic AI.
If you are rearchitecting your own pricing model, start with a conversation.
The full playbook is in Monetizing Agentic AI, available on Amazon in all three formats.