Intuit Inc.

Intuit makes the software that ordinary American businesses and households use to keep track of their money and to deal with the tax authorities. QuickBooks is the accounting system of millions of small businesses — the ledger where every invoice, bill, payroll run and bank transaction is recorded — and, increasingly, the place where those businesses take card payments, run payroll and borrow. TurboTax files about 39 million US federal tax returns a year, either as do-it-yourself software or with a human tax expert on video ("TurboTax Live"). Credit Karma shows consumers their credit scores free and earns a fee when they take out a card, loan or insurance policy it recommends. Mailchimp sends marketing emails for small businesses. A ten-year-old's version: Intuit keeps the scorebook for small shops and helps families fill in the government's homework — and gets paid a little every month, and a little every April.

It is an extraordinarily profitable business. In the year to July 2026 Intuit sold $21.4bn, up 14%; it kept 80% as gross margin, earned $5.9bn of operating profit after all costs including $2.1bn of stock-based pay, and generated $8.6bn of free cash flow. It spends almost nothing on factories or equipment. About 60% of revenue now comes from businesses (QuickBooks and its add-ons) and 40% from consumers (TurboTax, Credit Karma, professional tax software).

And yet the shares, at $325.15 on 16 October 2026, are 52% below their October 2025 high and 60% below the record of July 2025. In 2026 investors concluded that AI agents — software that can read a bank statement, categorise a transaction or fill in a tax form — would make Intuit's products unnecessary. The company then gave them some evidence: TurboTax lost customers to cheaper rivals, Intuit cut 17% of its staff, and it guided to slower growth. This report asks which of those fears is right.

Two lessons for an owner. The first is that Intuit has outlived its most dangerous rivals before. In the 1990s Microsoft tried first to beat Quicken with its own product and then to buy Intuit outright; Intuit kept its customers because its software was simpler and its users would not retrain. Thirty years later the threat is software that needs no training at all, which is why this report takes it seriously. The second is that the long record is excellent and the recent one is not. Over twenty years Intuit has returned 12.6% a year with dividends, beating the S&P 500; over ten, 12.8%, a little behind it; over five, it has lost 9% a year, because investors who bought in 2021 paid more than 50 times earnings for growth that did arrive — revenue has risen 68% since — but at a price that left no room for a scare.

QuickBooks: a subscription with services on top. A small business pays QuickBooks Online a monthly fee — list prices rose again in August 2026, with the mid-tier Plus plan reportedly going from $115 to $140 a month — and then pays transaction fees for card payments, a per-employee fee for payroll, and interest on QuickBooks Capital loans. The key measure is revenue per customer: in FY2026 online-ecosystem revenue per customer rose 15% while paying customers grew only 3%, to 8.9 million. That is pricing power — and a warning that most growth now comes from charging existing customers more rather than winning new ones. The mid-market (QuickBooks Advanced and the new Intuit Enterprise Suite, aimed at firms that would otherwise move to NetSuite) grew 39%.

TurboTax: a season, two products. Do-it-yourself TurboTax is software: high margins, price-sensitive customers, and competition from cheaper or free alternatives. TurboTax Live connects the filer with a credentialed expert — a service business with human costs, but one where customers pay several times more and value trust over price. In FY2026 Live revenue rose 37% to 53% of TurboTax revenue, while total federal units fell 2% to 39.0 million and free filers fell from 8 million to 7 million. The mix is moving up-market, deliberately.

Where the costs sit. The segment margins (77% for business, 73% for consumer) look enormous because Intuit books most of research and development, the human tax experts and much of marketing in "other corporate expenses" — $7.3bn in FY2026. After those, stock pay ($2.1bn) and the amortisation of acquisitions, the GAAP operating margin was 27.4%. On fiscal 2027 guidance it rises to about 32%.

Buffett's 1991 letter defines an economic franchise as a product that is needed or desired, "thought by its customers to have no close substitute", and not subject to price regulation — and says the proof is the ability to "regularly price its product or service aggressively" while earning high returns. QuickBooks passes that test as clearly as any software in America: Intuit has raised its prices year after year, and in 2026 revenue per customer rose 15% while customers kept coming. TurboTax passed it for twenty years and failed it at the low end in 2026, when management admitted it had lost "quality DIY customers to lower-cost providers" and would deliberately accept lower prices to win them back. His 2007 warning fits the AI question exactly: "A moat that must be continuously rebuilt will eventually be no moat at all." The ledger's moat does not need rebuilding each year; the DIY tax moat now does. We score the moat 7: wide where most of the profit is, narrowing where most of the headlines are.

Act one: the AI sell-off (January–April 2026). The shares fell from $662 at the end of 2025 to $362 on 9 April without a single disappointing Intuit result — the company beat and reiterated its guidance in November and February. They fell with all application software, on the fear that AI agents would make seat-based and form-filling software redundant. The two worst days line up with product launches by Anthropic: Claude Cowork plug-ins on 3 February (Intuit −10.9%) and Managed Agents on 9 April (−7.1%). The irony is that Intuit had signed a $100m-plus deal to put TurboTax, QuickBooks and Credit Karma inside ChatGPT in November 2025, and a partnership to run its tools inside Claude in February 2026.

Act two: the leak (May–August 2026). On 20 May Intuit reported a tax season in which TurboTax revenue grew 7% but online units fell, its share of e-filed returns slipped about a point and free customers fell from 8 million to 7 million — in a year when the IRS had shelved its free Direct File service and total self-prepared e-filings grew. The same evening it announced a 17% cut in its workforce. The shares fell 20% the next day, their worst day since March 2003, and reached a closing low of $255.07 on 25 June. In August the fiscal 2027 guidance — revenue +9–10%, TurboTax +2–3% — confirmed slower growth, and JPMorgan and Bank of America downgraded the stock.

Our reading: AI is eating the market's patience faster than it is eating Intuit. The business Intuit actually earns most of its money from — the ledger and the money services built on it — grew faster in FY2026 than at any time since the pandemic, and management's admission of trouble was about DIY tax, which is roughly an eighth of revenue. The TurboTax leak is real and deserves the attention it got: a franchise that lost customers in a year when its government competitor stepped aside has less pricing power than investors assumed. But it is a price problem at the bottom of the market, and Intuit has chosen to fix it by charging less there while selling more assisted service above it. The genuine long-run question — whether a future AI agent with access to a household's documents can file a return as reliably as TurboTax — is open; the evidence of 2026 is that it has not happened yet, and that Intuit intends to be the company whose agent does it. Goodarzi in August: "AI will be a disruptor, and we intend to be the disruptor." The next test is the first quarter of fiscal 2027, guided at revenue growth of 11%, reported in late November.

The pattern. Intuit's competitors split cleanly in two. In small-business accounting nobody has dented QuickBooks in the United States; Xero is strong abroad and the AI-native ledgers are still small and mostly build on top of QuickBooks. In consumer tax the bottom of the market has become a price war, and Intuit has decided not to fight it on price except selectively. The more interesting competitors are the AI companies Intuit has chosen to partner with: by putting its tools inside ChatGPT and Claude it gains distribution and risks becoming a supplier to whoever owns the customer's conversation. The comparison on our board is Adobe, another dominant application-software franchise priced in 2026 as if AI had already replaced it.

Integrity: no veto — with a scar. Intuit paid $141m to all fifty states in 2022 over the marketing of "free" TurboTax, and won its fight with the FTC on constitutional grounds in March 2026; its customer communications have not always been as straight as its accounts. Its financial reporting, on the other hand, just became more conservative by choice. Buffett asked in 1992: "If compensation isn't an expense, what is it?" Intuit has finally agreed. Capital allocation: mixed. Credit Karma worked; Mailchimp did not; a decade of buybacks barely kept the share count flat — but the 2026 buying, at around $300, is the best-priced buying Intuit has done in years. Owner mentality: weak at the top. The founder's trusts and the CEO sold about $350m between them near the peak, under pre-arranged plans; only one director has bought. We score management and capital 6.

Which earnings to believe. For years Intuit's preferred "non-GAAP" earnings excluded stock-based pay and the amortisation of acquired intangibles; in FY2026 they were $24.27 a share against GAAP earnings of $16.46. From fiscal 2027 stock pay is included, so the guided adjusted figure ($22.88–23.12) is lower than last year's while the business is growing faster than ever in profit terms — GAAP operating income is guided up 26–27%. The only remaining adjustment is the amortisation of acquired intangibles, a non-cash charge for deals already paid for. Our owner earnings sit in between: about $21 a share, roughly free cash flow without the tax catch-up, less stock pay, or adjusted earnings including stock pay.

Two things to watch in the cash. First, FY2026's $8.6bn of free cash flow includes about $1.28bn of lower cash taxes from the 2025 tax law that restored immediate expensing of US research spending — it will not recur at that size. Second, Intuit is now a lender of some size: it originated $6.8bn of QuickBooks Capital loans and tax-refund advances in FY2026, up from $4.0bn, and set aside $237m for credit losses. That is a sensible use of what it knows about its customers' cash flows, and a balance-sheet risk that pure software companies do not carry.

The verdict on the dividend: about as safe as dividends get, and too small to be the reason to own the shares. At $5.52 a year it yields 1.7% and takes about a sixth of free cash flow; there is no realistic scenario in which it is cut. What makes it interesting is its growth — fifteen straight annual raises averaging roughly 15% — and the fact that the board raised it by 15% in August, in the middle of the share-price collapse, layoffs and a growth reset. An investor buying at $325 gets a 1.7% yield that, on the past decade's pace, would double in five years. The comparison on our board is Microsoft, another low-yield, fast-raising dividend paid out of an enormous cash flow.

What is it worth? We discounted owner earnings of about $21 a share at 10% for ten years. If Intuit does roughly what it has guided — revenue growing high single digits to low double digits, margins widening as stock pay shrinks, and buybacks at today's prices — and owner earnings grow ~9% a year for five years and 5% thereafter, ending at 16 times, it is worth about $444. If AI agents and low-cost rivals squeeze both TurboTax and, eventually, QuickBooks — $19 growing 2% a year, 11 times — about $227, below the June low. If the platform keeps compounding as it did for twenty years — $23 growing 13% then 7%, 19 times — about $681, near last year's high.

At $325 the price sits about 27% below our central value, at the lowest multiple of earnings Intuit has traded on in more than a decade, for a business whose largest part is growing faster than ever and whose weakest part is being repaired. The pessimistic case is not fanciful — that is why the shares are here — but the price already pays for much of it. Our advice: begin accumulating, and add below $280, the September trading range, about 37% under our central value. This is not a stock to buy all at once: the next two tax seasons will show whether the TurboTax leak has been plugged, and AI news will move the price violently in both directions.