FY Guidance Hit-Rate
100%
Met-or-Beat (4/4 closed FYs)
Key Snapshot
$18.8B
FY25 Revenue (+16% YoY)
$21.0–21.2B
FY26 Revenue Guide (+12–13%)
$7.6B
FY25 Non-GAAP Op Inc
$8.6–8.7B
FY26 Op Inc Guide (+13–14%)
3M+
AI Agent Users (Q2 FY26)
85%+
Repeat Engagement on Agents
$6.1B
FY25 Free Cash Flow (+31%)
~55%
Stock Drawdown from 2024 Peak
Guidance vs. Actual — FY2022 to FY2026
| FY |
Metric |
Initial Guide (Aug) |
Final Guide |
Actual |
vs Initial |
Verdict |
| FY22 |
Revenue ($B) | 11.05–11.20 | 12.55–12.59 | 12.73 | +13.7% | Exceeded |
| GAAP EPS | 4.46–4.79 | ~$8.50 | $8.80 | +89% | Exceeded |
| Non-GAAP EPS | 11.05–11.45 | ~$11.85 | $11.96 | +5.6% | Exceeded |
| FY23 |
Revenue ($B) | 15.85–16.10 | 14.30–14.40 | 14.37 | −10.4% | Met (Lowered) |
| GAAP EPS | 9.05–9.50 | ~$10.20 | $10.45 | +13.5% | Exceeded |
| Non-GAAP EPS | 15.05–15.50 | ~$14.40 | $14.40 | −6.0% | Met (Lowered) |
| FY24 |
Revenue ($B) | 15.89–16.11 | 16.16–16.20 | 16.29 | +2.0% | Exceeded |
| GAAP EPS | 10.69–10.95 | ~$11.95 | $12.32 | +13.6% | Exceeded |
| Non-GAAP EPS | 16.17–16.47 | ~$16.80 | $16.92 | +3.4% | Exceeded |
| FY25 |
Revenue ($B) | 18.16–18.35 | 18.72–18.76 | 18.83 | +3.4% | Exceeded |
| GAAP EPS | 13.34–13.59 | ~$13.50 | $13.76 | +2.5% | Exceeded |
| Non-GAAP EPS | 19.16–19.36 | ~$20.10 | $20.18 | +5.3% | Exceeded |
| FY26 |
Revenue ($B) | 21.00–21.19 | 21.00–21.19 (reiterated Q2) | — | +12–13% YoY | Pending |
| GAAP EPS | 15.49–15.69 | 15.49–15.69 | — | +13–14% YoY | Pending |
| Non-GAAP EPS | 23.06–23.26 | 23.06–23.26 | — | +14–15% YoY | Pending |
Note: FY23 revenue initial guide assumed continued Credit Karma growth that was disrupted by the rate cycle; Intuit lowered the range mid-year (Q2 FY23) and beat the revised range. Every other FY beat both initial and final guidance ranges.
The AI Disruption Question — Addressed
Will AI kill Intuit?
Three of Intuit's product franchises — TurboTax, QuickBooks, Credit Karma — sit squarely in the path of generative AI. The bear case is straightforward: a $20/month LLM with tax-form OCR can do what TurboTax charges $90 to do; AI bookkeeping startups can replace QuickBooks workflows; LLM agents can route consumers to financial products without Credit Karma's intermediation. So far, the numbers are moving the other way: every reported quarter through Q2 FY26 has shown accelerating, not decelerating, growth — driven by the same AI that supposedly disrupts them. This section lays out the bear case, the bull counter, the metrics, and what's still unresolved.
🐻 Bear Case — AI Disrupts Intuit
1. TurboTax → Free LLM Tax Agents
Tax filing is structured-data extraction. Claude or ChatGPT with a W-2 PDF can produce a complete federal return in minutes. H&R Block already shipped "AI Tax Assist"; Cash App Taxes is free; commodity LLM access at $20/month undercuts TurboTax's $89–$209 paid tiers. Volume pressure on the DIY base could force pricing concessions.
2. QuickBooks → AI-Native Bookkeeping
Digits (autonomous general ledger, led by Xero co-founder Craig Walker), Pilot, Paro, and Puzzle automate 90%+ of SMB bookkeeping. A solo-preneur with Claude can categorize a year of bank CSVs and generate a P&L without any subscription. Net new SMB additions on QBO low-tier could be at risk first.
3. Credit Karma → Disintermediated by Agents
Credit Karma's moat is recommendations + lead generation to lenders. An LLM-powered personal finance agent can compare every credit card / loan / insurance product on the market without Credit Karma's carousel. Affiliate-style monetization is the most exposed business model in the company.
4. Margin Compression Risk
Even if Intuit wins the platform layer, embedding Claude / GPT into every workflow has a per-token cost the legacy SaaS model never had. If competitors race to free, gross margin could compress before agentic ARPU uplift fully materializes.
🐂 Bull Case — Intuit's AI Moat Holds
1. Domain Data + HI = "Intuit Operating System"
Intuit owns 30+ years of small-business GL data, 100M+ tax returns, and a network of 17,000+ tax pros and bookkeepers ("HI" = Human Intelligence). Frontier LLMs are a commodity input; the proprietary financial LLMs trained on this dataset, plus the human expert layer, are what's hard to replicate.
2. AI Agents Are Driving ARPU UP, Not Down
Q1 FY26: QBO Accounting revenue +25% YoY. QBO Advanced generates ~5x ARPC vs. QBO Core; payroll & payments penetration is 9–12 ppt higher in Advanced. 800K existing QBO customers qualify to upsell to Advanced or Intuit Enterprise Suite (mid-market). AI agents ("Intuit Assist") are the upsell hook.
3. Frontier-Lab Partnership De-Risks the Stack
Feb 2026: multi-year Anthropic deal — Intuit embeds Claude Agent SDK; QuickBooks, TurboTax, Credit Karma, Mailchimp expose MCP servers inside Claude. Intuit data & models stay proprietary inside Intuit's perimeter. This is the inverse of being disrupted: Intuit becomes the financial back-end Claude users reach for.
4. Operating Leverage Still Expanding
FY25 GAAP op income +36% to $4.9B. Q2 FY26 GAAP op income +44% YoY to $855M. FY26 guide implies non-GAAP op margin ~41% (vs 40% FY25). No margin compression yet; AI is automating cost out faster than it's compounding compute spend.
Sasan Goodarzi (CEO) on AI — His Own Words
"Over 3 million customers have leveraged agents to do the work for them, with all-time repeat engagement of more than 85%."
— Q2 FY26 Earnings Call · Feb 26, 2026
"Our disruptive AI-native mid-market platform is fueling the success of growing businesses, and we are further scaling our investment, product innovation, and go-to-market motions to accelerate customer adoption."
— Q2 FY26 Press Release · Feb 26, 2026
"Intuit's moat comes from proprietary data, domain-specific AI models including knowledge engineering and machine learning, Intuit financial LLMs, coupled with human expertise. Human Intelligence is a massive differentiator."
— Q2 FY26 Earnings Call · Feb 26, 2026
"Partnerships with LLM providers are structured to prevent proprietary data or AI models from leaving Intuit systems."
— Q2 FY26 Earnings Call · responding to a question on competitive AI risk
"Disruption in the assisted tax and mid-market is driven by AI and HI. Core is Intuit's proprietary data, domain-specific AI models, and human expertise. Context is everything else — and we are happy to partner on context."
— Q2 FY26 Earnings Call (Goodarzi's Core/Context framework for AI strategy)
The Frontier-Model Pivot — Intuit × Anthropic (Feb 2026)
Strategic Partnership
Announced February 2026. Intuit and Anthropic struck a multi-year deal to embed Claude across Intuit's product suite and to expose Intuit's products as MCP-callable tools inside Claude.
Two-way integration:
- Inside Intuit: Claude Agent SDK powers next-gen "Done-for-You" agents in TurboTax, QuickBooks, Credit Karma, and Mailchimp. Intuit's proprietary data and financial LLMs stay inside Intuit's environment.
- Inside Claude: MCP servers let Claude users invoke TurboTax (file taxes), QuickBooks (open ledger, run reports), Credit Karma (compare offers), and Mailchimp (send campaigns) directly. Intuit becomes the financial back-end that frontier-model agents reach for.
Why it matters: the deal directly answers the bear case. If a Claude user asks "do my taxes" — the answer is TurboTax via MCP. If they ask "show me my P&L" — it's QuickBooks. The disruption vector becomes the distribution vector.
Are the Metrics Backing Goodarzi Up?
+25%
Q1 FY26 YoY
QuickBooks Online Accounting Revenue
+23%
Q2 FY26 YoY
Credit Karma Revenue
+12%
Q2 FY26 YoY
TurboTax Live (Assisted) Revenue
~5x
vs QBO Core ARPC
QBO Advanced ARPC Multiple
800K
Upsell pipeline
QBO Customers Eligible for Advanced/IES
+44%
Q2 FY26 YoY
GAAP Operating Income
40.3%
→ ~41% guide FY26
FY25 Non-GAAP Operating Margin
$6.1B
+31% YoY
FY25 Free Cash Flow
What's Still Unresolved
- Free-tier consumer erosion. Cash App Taxes (free) + IRS Direct File (now in 25+ states) + LLM-assisted DIY filing could compress TurboTax's free-to-paid funnel. FY26 Consumer guide of +8–9% is still constructive but worth watching every Q3 (tax season) print.
- SMB micro-tier churn to AI-natives. Sub-$20/month bookkeeping startups (Digits, Pilot Lite) target the same 1–5 employee cohort that QBO Self-Employed and QBO Simple Start serve. Net QBO subscriber adds vs. ARPU uplift is the key disclosure to watch.
- Compute-cost curve. Even with Anthropic partnership, Intuit Assist's per-interaction cost rises as more customers move from form-fill to agent-driven workflows. Margin trajectory is positive today; the question is whether it scales when tax season puts millions of agents in flight at once.
- Stock has priced in heavy disruption. Down ~55% from the 2024 peak — sentiment has already discounted significant erosion. That's a setup for outperformance if Goodarzi's "AI is accretive" thesis keeps printing each quarter.