Intuit Just Stopped Being a Software Company

TL;DR

Intuit cut 17% of its workforce (3,000 people) during a 48% profit jump and signed multi-year deals with Anthropic and OpenAI to license its tax and financial data to Claude and ChatGPT. The story is a category change: Intuit is moving from a software company that serves users to a data company that licenses outcomes to foundation models. Every business with proprietary data faces the same recategorization. The strategic move for leaders is to inventory the unique data their company owns and decide whether to build, license, or sell access to it.

---

Intuit cut 3,000 people this week. About 17% of its workforce. The headlines called it a layoff. The headlines are wrong.

What actually happened at Intuit is a category change. The company that built TurboTax, QuickBooks, Credit Karma, and Mailchimp is no longer a software business. It is becoming a data business that licenses tax and financial outcomes to foundation models. The 3,000 people who got cut are casualties of that recategorization. The people who remain are about to learn what their actual product is.

This matters far beyond Intuit. If you sit on proprietary data of any kind, your category is about to change too.

What the Memo Actually Said

CEO Sasan Goodarzi sent an internal memo on May 20 announcing the cuts. TechCrunch published the full context. The framing was clean. Simplify the corporate structure. Reduce complexity. Deliver better AI products across the whole portfolio.

What did not appear in the early coverage was the strategic move underneath the memo. Intuit signed multi-year partnerships with both Anthropic and OpenAI. The arrangement is specifically structured to feed Intuit's tax, accounting, and credit data into Claude and ChatGPT.

Read that carefully. Intuit is not just integrating these models into TurboTax. Intuit is making its data the substrate that other people's AI products run on. That is the move of a data licensing company. It is not the move of a software company.

The Financials Make the Move Look Inevitable

In its most recent quarter, Intuit reported $4.65 billion in revenue, up 17%, and net profit of $693 million, a 48% jump. Numbers that good rarely produce 17% layoffs. They did this time because Goodarzi looked at where the next decade of value is being captured and concluded that it is not in software interfaces.

The interface era of fintech is closing. Twenty years ago, the moat was the workflow inside TurboTax. The questions you answered. The hand-holding. The user experience that walked a small business owner through quarterly estimates.

That moat is now an AI feature. Any general-purpose model can walk a user through the same questions, given the right data. So Intuit is racing to make sure that the "right data" only comes from Intuit. The data becomes the moat. The interface becomes a commodity.

Why This Pattern Will Repeat at Every Data-Rich Company

If you work at a healthcare company, a logistics company, an insurance company, a legal firm, or any business that has accumulated decades of operational data, the Intuit restructuring is a preview of the next 24 months for you.

The pattern works like this. Your company has data that nobody else has. Foundation models are hungry for high-quality, domain-specific data. You can either build worse AI products on top of your own data, or license your data to better AI products built by others, or do both. The third option produces the highest revenue per dollar of human cost.

In every case, the staff count goes down. Because the data, once licensed, no longer needs hundreds of human analysts to interpret it. The model does that, at scale, for the customers paying to use it.

The companies running this playbook well include Bloomberg, which built a domain-specific model on its own financial corpus. Reuters is doing the equivalent in legal. Epic is positioned for it in healthcare. And now Intuit is doing it in personal finance and small business accounting.

The execs who win in this transition are the ones who can answer two questions cleanly. What data do we own that nobody else does. And what does it look like to monetize that data as a service rather than as a feature.

What the Layoffs Tell You About Where the Cuts Land

In any data-licensing transition, three groups of employees become redundant fastest. Customer support staff, because foundation models can answer most user questions natively. Junior product managers, because feature roadmaps shrink when most features are model capabilities. And the layer of analysts who used to translate raw data into dashboards, because models do that translation at a fraction of the cost.

The three groups that get more valuable are data engineers, agent designers, and trust and safety specialists who keep the licensed data clean and compliant.

This is the Hive Structure showing up inside a single company. Bees are agents. The beekeepers are the people who ensure the bees produce safe, accurate, monetizable outputs. The middle layer is gone.

The Token Economy Is Now Selling Tax Software

I wrote about the token economy in How to Do More with Less Using AI. The argument was that the unit of organizational production is moving from labor hours to compute tokens, and CFOs would eventually have to report performance in tokens per outcome.

Intuit's move accelerates this. When you license your data to model providers, the revenue you collect from that licensing is denominated in usage. Tokens consumed. Calls to the model. Outputs produced. Your revenue per employee figure starts to look meaningless. Your revenue per token consumed becomes the metric that explains the business.

When that happens at Intuit, every other vertical software company with proprietary data is going to feel pressure from its board to follow. Salesforce data. ServiceNow data. Workday data. Each of these companies is sitting on the same opportunity. Each will face the same question. License it, or watch a competitor license its data and price you out.

What Small and Mid-Sized Companies Should Do Right Now

If you don't run a multi-billion dollar SaaS company, the Intuit lesson still applies. Just at a smaller scale.

Your competitive advantage in the agent era is the data you own that nobody else does. Your customer notes. Your service histories. Your billing patterns. Your operational telemetry. Most companies have far more proprietary data than they realize, and almost none of them have a strategy for turning that data into AI leverage.

Three steps for this week.

First, run a data inventory. Not a list of databases. A list of operational decisions your company makes well because of data you uniquely hold. Each of those is a candidate for AI monetization or AI-driven cost reduction.

Second, ask whether any external partner could replicate those decisions if they had your data. If yes, that data is leverage. If no, you have a domain advantage that AI can amplify but not replace.

Third, decide whether you are building, licensing, or selling. The three paths produce different org charts. The Intuit memo this week shows what the licensing path looks like when it is run at scale.

The Test

Pull up the Intuit press release for the layoffs and the Anthropic and OpenAI deals. Read them in the same sitting. Notice the language. The layoffs are framed as a refocus. The deals are framed as a strategic acceleration. Both are aspects of the same decision.

Then ask the same question of your own company. Where is your refocus, and where is your acceleration? If you cannot identify both clearly, you are still operating in the software era while your competitors are operating in the data era.

What category is your business actually in this year?

Sharon Gai is an AI transformation strategist, keynote speaker, and author of How to Do More with Less Using AI. She advises Fortune 500 companies on AI adoption and organizational redesign.