# How Figma makes money

CXProWealth Business Lens · Figma (FIG) · Q2 FY2026

Read from the 10-Q for the period 2026-04-01 to 2026-06-30, filed 2026-08-05 with the SEC (accession 0001628280-26-053348): https://www.sec.gov/Archives/edgar/data/1579878/000162828026053348/

Business Lens explains how a company earns, spends and keeps its money, from its own filings. It is insight and context, never a view on the shares, a target, or advice. Figures are raw units as filed; percentages are percentage points.

## Headline figures

- Revenue: $370.1M (+48.2% on the year)
- Operating income: -$117.3M (operating margin -31.7%)
- Net income: -$112.2M (net margin -30.3%)
- Gross margin: 83.7%
- Free cash flow: $54.2M (14.6% of revenue)
- Cash conversion (operating cash flow / net income): -0.54x
- Capital expenditure: 1.8% of revenue

## Where the money goes (Q2 FY2026)

- Revenue → Cost of revenue: $60.5M
- Revenue → Gross profit: $309.6M
- Gross profit → Total operating expenses: $309.6M
- Total operating expenses → Research and development: $167.3M
- Total operating expenses → Sales and marketing: $154.9M
- Total operating expenses → General and administrative: $104.7M
- Income (loss) from operations (loss) → Total operating expenses: $117.3M

## Income statement, as filed

| Line | Q2 FY2026 | Of revenue |
|---|---:|---:|
| Revenue | $370.1M | 100% |
| Cost of revenue | $60.5M | 16.3% |
| **Gross profit** | $309.6M | 83.7% |
| Research and development | $167.3M | 45.2% |
| Sales and marketing | $154.9M | 41.8% |
| General and administrative | $104.7M | 28.3% |
| **Total operating expenses** | $426.9M | 115.4% |
| **Income (loss) from operations** | -$117.3M | 31.7% |
| Other income, net | $7.6M | 2.1% |
| **Income (loss) before income taxes** | -$109.7M | 29.6% |
| Provision for income taxes | $2.5M | 0.7% |
| **Net income (loss)** | -$112.2M | 30.3% |
| Less: net income attributable to participating securities | ($0) | 0.0% |
| **Net income (loss) attributable to common stockholders** | -$112.2M | 30.3% |
| Net income (loss) per share, basic (in usd per share) | -$0 | 0.0% |
| Net income (loss) per share, diluted (in usd per share) | -$0 | 0.0% |
| Weighted-average shares outstanding used in computing net income (loss) per share attributable to common stockholders, basic (in shares) | 527,460,000 |  |
| Weighted-average shares outstanding used in computing net income (loss) per share, diluted (in shares) | 527,460,000 |  |

## History

| Period | Revenue | Operating income | Net income | Operating cash flow | Capital expenditure | Free cash flow |
|---|---:|---:|---:|---:|---:|---:|
| FY2023 | $504.9M | -$73.5M | $737.8M | $1.0B | $3.7M | $1.0B |
| FY2024 | $749.0M | -$877.4M | -$732.1M | -$61.7M | $2.0M | -$63.7M |
| FY2025 | $1.1B | -$1.3B | -$1.3B | $250.7M | $4.4M | $246.2M |

Latest trend (the newest full year against the year before; a reading, not a forecast):

- Revenue: +41.0% (year before +48.4%): slowing
- Operating income: — (year before —): too short a history to read
- Net income: — (year before -199.2%): too short a history to read
- Free cash flow: -286.6% (year before -106.1%): slowing

## Operating figures stated in the filing

- Paid Customers with more than $10,000 in ARR: 15,964 customers (as of 2026-06-30). "Paid Customers with more than $10,000 in ARR 15,964 11,906" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)
- Paid Customers with more than $10,000 in ARR: 11,906 customers (as of 2025-06-30). "Paid Customers with more than $10,000 in ARR 15,964 11,906" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)
- Paid Customers with more than $100,000 in ARR: 1,635 customers (as of 2026-06-30). "Paid Customers with more than $100,000 in ARR 1,635 1,119" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)
- Paid Customers with more than $100,000 in ARR: 1,119 customers (as of 2025-06-30). "Paid Customers with more than $100,000 in ARR 1,635 1,119" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)
- Net Dollar Retention Rate: 136.0% (as of 2026-06-30). "Net Dollar Retention Rate 136 % 129 %" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)
- Net Dollar Retention Rate: 129.0% (as of 2025-06-30). "Net Dollar Retention Rate 136 % 129 %" (Item 2. Management’s Discussion And Analysis Of Financial Condition And Results Of Operations)

## Key insights

### Large accounts are expanding faster than the customer base

Net dollar retention rose to 136% from 129%, and accounts above one hundred thousand dollars are now 10.2% of larger customers, up from 9.4%.

Revenue rose 48.2%, to $370.1M from $249.6M. The filing says the increase was "primarily driven by growth and expansion in total Paid Customers". Customers above ten thousand dollars of annual recurring revenue rose to 15,964 from 11,906, and those above one hundred thousand dollars rose to 1,635 from 1,119. _(from the filing)_
Net dollar retention, measured on customers above ten thousand dollars of annual recurring revenue, was 136% against 129% a year earlier. After churn and contraction, the same group of accounts paid more than it did twelve months before, and the pace of that expansion went up rather than down. _(from the filing)_
The mix is moving up. Accounts above one hundred thousand dollars made up 10.2% of the larger-customer group, against 9.4% a year earlier. Growth is coming from organizations rolling Figma out more widely, which is what the multi-product seats and enterprise tiers are built to capture. _(analysis)_
The filing expects net dollar retention to fluctuate or decline as the revenue base grows and penetration rises, and says its pricing changes make comparisons with earlier periods less meaningful. _(from the filing)_

### AI hosting pulled gross margin down, then it partly recovered

Cost of revenue grew 116.8% against revenue growth of 48.2%, cutting gross margin to 83.7% from 88.8%, though it was 79.4% in the March quarter.

Cost of revenue was $60.5M against $27.9M a year earlier, 16.3% of revenue against 11.2%. The filing names the cause: higher technical infrastructure and hosting costs relating to AI and to more usage by paid users. Gross profit grew 39.6%, slower than revenue. _(from the filing)_
Within the year the direction turned. In the March quarter cost of revenue was $68.7M on revenue of $333.4M, a gross margin of 79.4%. In the June quarter cost of revenue fell to $60.5M while revenue rose to $370.1M. March is when Figma began enforcing AI credit limits, and the filing says AI costs may move as it makes changes "intended to manage usage and optimize the efficiency of our AI-related spend". _(from the filing)_
The timing fits metering doing its job, but the filing does not attribute the recovery to the credit limits, and one quarter is not a trend. Gross margin also understates total AI cost: hosting and AI for free users is booked in sales and marketing, not in cost of revenue. _(analysis)_
The company expects its AI investments to "negatively impact our gross margins and operating margins" in the short term, and calls the longer-term impact "currently unknown". _(from the filing)_

### A $117.3M operating loss alongside $54.2M of free cash

Operating expenses reached 137.9% of gross profit, mostly from post-IPO stock pay the filing names, yet the quarter still produced a 14.6% free cash flow margin.

Operating expenses rose 94.3%, to $426.9M, or 115.4% of revenue against 88.0% a year earlier. General and administrative rose 169.0% and research and development 101.5%. The filing attributes most of the increase in each line to stock-based compensation and related payroll taxes recognized after and as a result of the initial public offering. _(from the filing)_
Cash moved the other way. Operating cash flow was $60.9M and free cash flow $54.2M, a 14.6% margin. Subscriptions are generally invoiced in advance, and capital spending of $6.7M was 1.8% of revenue. Over the last full year the same split held: free cash flow of $246.2M against an operating loss of $1.29B. _(from the filing)_
Free cash flow did not keep pace with revenue. It was $61.3M a year earlier, a 24.6% margin, so cash generation fell while revenue grew 48.2%. _(from the filing)_
Stock pay uses no cash, but shareholders bear it through dilution, so free cash flow on its own overstates the economics. The reported loss narrows only if revenue grows faster than that expense. _(analysis)_

### The year-ago profit came almost entirely from investment income

A year earlier other income was $37.0M, or 94.7% of the $39.1M pre-tax profit, while operations earned just $2.08M.

In the June quarter a year earlier, income from operations was $2.08M, other income $37.0M, pre-tax income $39.1M and net income $28.2M. Other income consists mainly of interest on cash and securities and gains or losses on equity holdings, including a Bitcoin exchange traded fund. _(from the filing)_
This quarter other income fell to $7.6M. The filing attributes the decrease to an unfavorable change in the fair value of equity securities, primarily the Bitcoin fund moving from an unrealized gain a year earlier to an unrealized loss. Net income of $28.2M became a net loss of $112.2M. _(from the filing)_
The swing in net income mixes two separate stories: an operating line driven by equity pay, and investment results that move with Bitcoin. Neither says much about demand, which the customer counts and retention track more directly. _(analysis)_

### Competitive edge

- **Free users carry Figma into paying organizations** Hosting and AI for free users are booked in sales and marketing, 41.8% of revenue, treating the free plan as the main acquisition channel.
- **Customers pay up front and the servers are rented** Capital spending was 1.8% of revenue and 11.0% of operating cash flow, so nearly all operating cash became free cash flow.
- **Each new product reaches another role in the same account** Research and development took 45.2% of revenue, up from 33.3%, funding products aimed at developers, product managers and marketers inside existing customers.

### What to watch

- **AI agents could do the work of paid seats** The filing warns its seat-based model may fit less well as AI and agents let fewer people produce the same output.
- **AI features may cost more than they earn** Figma depends on a few model providers and says price pressure could force it to run some AI features at low or negative gross margins.
- **Retention is expected to fluctuate as pricing keeps changing** The filing expects net dollar retention, now 136%, to fluctuate or decline, and says pricing changes make past comparisons less meaningful.

## Investor view

**AI now has a cost line at Figma: gross margin fell 5.2 points from a year earlier as revenue kept growing**

### How it makes money

Figma sells subscriptions to a browser-based platform where teams design, prototype and now build digital products. Contracts run monthly or annually, are generally non-cancelable, and "we typically invoice our customers in advance", recognizing revenue over the term. A free Starter plan brings people in; Professional, Organization and Enterprise plans carry the paid features. That produced $370.1M of revenue in the June quarter, against $249.6M a year earlier. _(from the filing)_
Delivering the software costs little next to what it sells for. Gross profit was $309.6M, a gross margin of 83.7%. Cost of revenue is mostly hosting and technical infrastructure, now including AI inference, plus payment processing and support for paid users. The heavier spending sits below that line: research and development took 45.2% of revenue, sales and marketing 41.8% and general and administrative 28.3%. _(from the filing)_
Growth comes mostly from customers spending more. The filing says the revenue increase was "primarily driven by growth and expansion in total Paid Customers". Paid customers above ten thousand dollars of annual recurring revenue rose to 15,964 from 11,906, and those above one hundred thousand dollars to 1,635 from 1,119. Net dollar retention in the larger group was 136%, against 129% a year earlier: the same customers pay more than they did twelve months before. _(from the filing)_
A second meter now sits on top of the seat. AI credits come with every Figma seat, and since March the company enforces the credit limits, selling monthly credit add-ons or pay-as-you-go usage beyond them. The free plan is treated as an acquisition cost: hosting and AI for free users is booked in sales and marketing, not in cost of revenue, so the gross margin reflects the cost of serving paying users. _(from the filing)_

### What shifted

Revenue rose 48.2%, to $370.1M from $249.6M, and climbed every quarter in between: $274.2M in the September quarter, $303.8M in December and $333.4M in March. The filing ties the gain to more paid customers and to existing ones buying more. _(from the filing)_
Cost of revenue grew 116.8%, to $60.5M, more than twice as fast as revenue. It now takes 16.3% of revenue, 5.2 points more than a year earlier, and gross margin fell to 83.7% from 88.8%. The filing names the cause: higher technical infrastructure and hosting costs relating to AI and to more usage by paid users. The trend runs back further: over the last full year cost of revenue was $185.5M, more than double the $87.5M of the year before, while revenue rose to $1.06B from $749.0M. _(from the filing)_
The margin turned up within the year. Gross margin was 79.4% in the March quarter and 83.7% in June, and cost of revenue fell to $60.5M from $68.7M while revenue rose. March is also when Figma began enforcing AI credit limits, and the filing says AI costs may move as it makes "operational and platform changes intended to manage usage and optimize the efficiency of our AI-related spend". _(from the filing)_
Operating expenses rose 94.3%, to $426.9M from $219.7M, or 115.4% of revenue against 88.0%. General and administrative rose 169.0% and research and development 101.5%. The filing attributes most of the increase in each line to stock-based compensation and related payroll taxes recognized after the initial public offering, with hiring and AI hosting adding to research and development. _(from the filing)_
The result swung from a $2.08M operating profit to a $117.3M operating loss, an operating margin of -31.7%. Net income of $28.2M became a net loss of $112.2M, with other income falling to $7.6M from $37.0M as a Bitcoin fund holding moved from gain to loss. Cash did not follow the loss: free cash flow was $54.2M, a 14.6% margin, against $61.3M and 24.6% a year earlier. _(from the filing)_

### Risks in the filing

AI could shrink the seat base. The company says customers may cut design, product and engineering headcount as AI does more of the work, and that agents working through its MCP server or through a single user's login could reduce the paid seats needed. In its words, "Our seat-based subscription model may become less aligned with customer value creation as AI automates or consolidates work that historically required multiple users." _(from the filing)_
AI may be costly to deliver and hard to charge for. Figma depends on "a limited number of third-party model providers" that may also compete with it, and pricing pressure could force it to "operate certain AI features at lower or negative gross margins for extended periods". After credit limits took effect in March it saw "elevated customer support volume, instances of customer dissatisfaction expressed through public and social channels, and reduced usage by certain customers". _(from the filing)_
Billing on credits and usage is harder to run than billing on seats. The filing lists invoice disputes, refunds, judgment calls on recognizing revenue from unused credits, collection risk and consumer-protection claims. It says the pricing changes make results harder to forecast and comparisons with earlier periods less meaningful, and it expects net dollar retention to fluctuate or decline. _(from the filing)_
Growth may slow as the business matures and as sales shift to larger buyers. Much past growth came from word-of-mouth adoption inside organizations; enterprise approval processes and admin controls on seat upgrades could slow that and raise the cost of winning customers. Competitors include AI-native and design-to-code tools, and holdings of Bitcoin and a Bitcoin fund add swings to reported results. _(from the filing)_
Of these risks, the cost of AI is already in the numbers, in the fall in gross margin from a year earlier, and the Bitcoin exposure shows in other income. The seat-replacement risk does not show yet: net dollar retention and the count of larger customers both rose. That is where the filing's warnings run furthest ahead of the reported figures. _(analysis)_

## Founder view: mechanisms

### Free users pull in paying teams

Individuals start free, invite colleagues, and the work spreads inside an organization before anyone signs a large contract. Figma then sells that organization the controls, security and tiers a company-wide rollout needs. The product carries much of the acquisition work, and the sales team arrives to expand accounts that already use it rather than to open cold ones. _(analysis)_

### Meter the AI, keep the seat

Each seat carries an allowance of AI credits, and usage beyond it is sold as add-ons or pay-as-you-go. The seat keeps recurring revenue predictable, and the meter ties the variable cost of inference to revenue from the people who drive it. Without the meter, the heaviest AI users would erode the margin on every seat. _(analysis)_

### Billed ahead, run on rented servers

Subscriptions are invoiced in advance and the product runs on a hosting partner's infrastructure, so customers fund operations before the service is delivered and the company buys almost no equipment. Much of the new pay is in stock rather than cash. Together these let a reported loss and positive cash flow sit side by side. _(analysis)_

### More products sold into the same account

Figma started as one design tool and has added whiteboarding, developer handoff, slides, prompt-to-prototype, website publishing, marketing assets and drawing, sold through multi-product seats. Each product reaches another role in the same company: product managers, developers, marketers. The account grows by covering more people and more steps of the work, not only by adding designers. _(analysis)_

## Member research

Signed-in members read, on the page: what Figma's numbers depend on from here, and the playbook behind each mechanism (what needs scale, what transfers, how to test it, the metric that would show it).

## Sources

- Figma 10-Q for the period to 2026-06-30: https://www.sec.gov/Archives/edgar/data/1579878/000162828026053348/ (2026-08-05)

Methodology business-lens-1.0 · schema 1.2.0 · run business-lens-2026-11-01 · as of 2026-11-01