Layer 8 of 8 · 27 companies
Software and applications: the return
Is the money coming back up the stack yet?
Every layer below is paid by the one above it; this layer is where the money has to come from. The software companies we track earned $193B over their latest four quarters, +37% in two years, while the buyers below them spent $566B on capex. Is the AI spending showing up in what customers pay for software?
- combined revenue of the software companies we track, latest four quarters
- $193B
- combined revenue of the software companies we track, latest four quarters
- their revenue change over two years
- +37%
- their revenue change over two years
- combined operating margin
- 18%
- combined operating margin
- Palantir’s revenue growth, the fastest in the group
- +79%
- Palantir’s revenue growth, the fastest in the group
The return has to come from somewhere
The money flowing down the stack is spending; the money that has to flow back up is revenue from people and businesses paying for AI. Some of that is visible in cloud segments, where AI labs and enterprises rent capacity. But for the build-out to pay off over time, AI has to make software more valuable to the companies that buy it: more seats sold, higher prices, new products, or software that replaces work customers used to pay people to do.
So this page asks a simple question of the listed software companies: is their growth accelerating as AI arrives? In aggregate, not by much yet. The group we track grew revenue +37% over two years, from $141B to $193B, at a steady gross margin of about 78%. That is healthy growth, but it is not the step change seen in chips or memory, and it is small next to the capex being spent on the layers below.
Quarterly revenue: software and applications
The layer’s 27 US-listed companies, five largest shown separately; Other is Palo Alto Networks, Workday, AppLovin, Atlassian, Palantir, Twilio, Snowflake, CrowdStrike, Datadog, HubSpot, Zscaler, Okta, MongoDB, Dynatrace, Elastic, Pegasystems, UiPath, Teradata, CCC Intelligent Solutions, GitLab, C3.ai, SoundHound AI
- Salesforce
- Adobe
- Intuit
- ServiceNow
- Shopify
- Other
Source: company 10-K and 10-Q filings (SEC XBRL); fiscal quarters aligned to the nearest calendar quarter. Companies appear from their first reported quarter, so newly listed ones (such as CoreWeave, Sandisk or Astera Labs) add to the total partway through.
Show the numbers
| Quarter | Salesforce | Adobe | Intuit | ServiceNow | Shopify | Other |
|---|---|---|---|---|---|---|
| Q1 2022 | $7.4B | $4.3B | $5.6B | $1.7B | — | $10.1B |
| Q2 2022 | $7.7B | $4.4B | $2.4B | $1.8B | — | $10.8B |
| Q3 2022 | $7.8B | $4.4B | $2.6B | $1.8B | — | $11.2B |
| Q4 2022 | $8.4B | $4.5B | $3.0B | $1.9B | — | $12.0B |
| Q1 2023 | $8.2B | $4.7B | $6.0B | $2.1B | — | $12.3B |
| Q2 2023 | $8.6B | $4.8B | $2.7B | $2.1B | — | $13.0B |
| Q3 2023 | $8.7B | $4.9B | $3.0B | $2.3B | — | $13.5B |
| Q4 2023 | $9.3B | $5.0B | $3.4B | $2.4B | — | $13.1B |
| Q1 2024 | $9.1B | $5.2B | $6.7B | $2.6B | $1.9B | $14.8B |
| Q2 2024 | $9.3B | $5.3B | $3.2B | $2.6B | $2.0B | $15.1B |
| Q3 2024 | $9.4B | $5.4B | $3.3B | $2.8B | $2.2B | $15.8B |
| Q4 2024 | $10.0B | $5.6B | $4.0B | $3.0B | $2.8B | $16.9B |
| Q1 2025 | $9.8B | $5.7B | $7.8B | $3.1B | $2.4B | $17.3B |
| Q2 2025 | $10.2B | $5.9B | $3.8B | $3.2B | $2.7B | $18.4B |
| Q3 2025 | $10.3B | $6.0B | $3.9B | $3.4B | $2.8B | $19.3B |
| Q4 2025 | $11.2B | $6.2B | $4.7B | $3.6B | $3.7B | $20.8B |
| Q1 2026 | $11.1B | $6.4B | $8.6B | $3.8B | $3.2B | $22.1B |
| Q2 2026 | $11.3B | $6.6B | $4.4B | $4.0B | $3.6B | $23.7B |
Where AI is showing up
Growth is concentrated. Over their latest four quarters the fastest growers were Palantir (+79%), AppLovin (+61%), SoundHound AI (+55%) and Shopify (+33%). Palantir sells data and AI platforms to governments and enterprises and runs at a 43% operating margin. AppLovin uses machine-learning models to place mobile ads, and its 77% operating margin is the clearest example in the group of AI turning directly into profit. SoundHound AI, the smallest company in the group, grew largely by buying other companies. The data platforms that AI workloads read from, Snowflake and Datadog, also grew faster than the group.
At the other end, among the largest companies, Salesforce (+11%), Adobe (+12%) and Workday (+13%) grew most slowly. These are established companies selling AI features into an existing customer base. AI may be protecting their position more than expanding it, and the reported numbers don’t yet show acceleration. Among the smaller ones, C3.ai (−38%), Pegasystems (+1%) and Teradata (+1%) barely grew or shrank. C3.ai, whose whole business is AI software, is a reminder that selling AI is not the same as growing with it.
The largest AI software revenue isn’t in this group at all. OpenAI and Anthropic are private, and the clearest window onto their scale is in their partners’ filings: Microsoft reported $24.1 billion of revenue from commercial arrangements with OpenAI in its latest fiscal year, and Amazon reports Anthropic’s expanded commitment to AWS. Much of what those labs earn from users flows back down the stack as cloud spending.
Revenue growth by company, latest four quarters vs the four before
Change in revenue from the year before
Source: company 10-K and 10-Q filings (SEC XBRL); fiscal quarters aligned to the nearest calendar quarter.
Show the numbers
| Company | Revenue | Change | Gross margin |
|---|---|---|---|
| Palantir | $6.2B | +79% | 85% |
| AppLovin | $6.8B | +61% | 88% |
| SoundHound AI | $203M | +55% | 42% |
| Shopify | $13.3B | +33% | 48% |
| Snowflake | $5.4B | +32% | 67% |
| Datadog | $4.0B | +32% | 80% |
| Atlassian | $6.6B | +26% | 85% |
| MongoDB | $2.8B | +25% | 73% |
| Zscaler | $3.4B | +25% | 77% |
| Palo Alto Networks | $11.5B | +24% | 70% |
| CrowdStrike | $5.4B | +24% | 75% |
| GitLab | $1.1B | +23% | 86% |
| ServiceNow | $15.4B | +22% | 73% |
| HubSpot | $3.4B | +21% | 83% |
| Dynatrace | $2.1B | +18% | 81% |
| Twilio | $5.6B | +18% | 49% |
| Elastic | $1.8B | +16% | 76% |
| UiPath | $1.7B | +15% | 83% |
| Intuit | $21.4B | +14% | — |
| Workday | $10.2B | +13% | — |
| Adobe | $26.0B | +12% | 89% |
| Salesforce | $43.9B | +11% | 77% |
| Okta | $3.1B | +11% | 78% |
| CCC Intelligent Solutions | $1.1B | +11% | 74% |
| Teradata | $1.7B | +1% | 61% |
| Pegasystems | $1.7B | +1% | 76% |
| C3.ai | $232M | −38% | 29% |
The risk to software from AI itself
AI cuts both ways for this layer. It can make software more valuable, but it can also lower the cost of building software, let customers replace some applications with AI agents, and shift pricing from seats (which shrink if AI does the work) to usage. For companies that charge per user, an AI that makes each user more productive may mean fewer users.
That is why the reported numbers matter more than the announcements. A company whose AI products are working should show it in at least one of three places: faster revenue growth, rising gross or operating margins as AI lowers its own costs, or larger backlog as customers sign bigger, longer contracts. Our AI Signal Gauge is built to test exactly that from the filings, and its public evidence pages show what it reads for several of these companies.
How this layer closes the loop
The buyers’ capex of $566B over four quarters compares with $193B of revenue for the whole software group we track. The comparison is rough (most AI spending is paid back through cloud rental, consumer subscriptions and advertising, not only enterprise software), but it captures the scale of what has to happen. For the stack to earn a return, revenue at the top has to keep compounding for years after the spending peaks.
The encouraging signs are the accelerating cloud segments and the few software companies turning AI into visible growth and margin. The open question is whether that spreads from a handful of winners to the broad base of software customers. That is the single most important thing to watch across the whole AI stack.
The companies, side by side
Latest four reported quarters for US filers; latest fiscal year for 20-F filers (marked). Underlined names have their own company page; the others link to the company’s filings on SEC EDGAR. Margins are left blank where a company doesn’t report gross profit or cost of revenue.
| Company | Revenue | vs year before | Gross margin | Operating margin | Free cash flow | Backlog (RPO) |
|---|---|---|---|---|---|---|
| Salesforce CRMCRM and Agentforce agents | $43.9B | +11% | 77% | 20% | $15.2B | $66.3B |
| Adobe ADBECreative and document software with generative features | $26.0B | +12% | 89% | 36% | $10.6B | $22.2B |
| Intuit INTUTax and accounting software with AI assistants | $21.4B | +14% | — | 27% | $8.7B | — |
| ServiceNow NOWWorkflow platform with AI agents | $15.4B | +22% | 73% | 10% | $4.8B | $29.7B |
| Shopify SHOPCommerce platform | $13.3B | +33% | 48% | 14% | $2.4B | — |
| Palo Alto Networks PANWNetwork and cloud security | $11.5B | +24% | 70% | 6% | $4.1B | $21.2B |
| Workday WDAYHR and finance software | $10.2B | +13% | — | 11% | $2.8B | — |
| AppLovin APPAd platform driven by machine-learning models | $6.8B | +61% | 88% | 77% | — | — |
| Atlassian TEAMCollaboration and developer tools | $6.6B | +26% | 85% | 0% | $1.3B | $4.8B |
| Palantir PLTRData and AI platforms for government and enterprise | $6.2B | +79% | 85% | 43% | $3.4B | $4.9B |
| Twilio TWLOMessaging and voice APIs that AI agents talk through | $5.6B | +18% | 49% | 5% | — | $138M |
| Snowflake SNOWData cloud that AI workloads read from | $5.4B | +32% | 67% | −23% | $1.2B | $9.0B |
| CrowdStrike CRWDEndpoint and cloud security | $5.4B | +24% | 75% | −2% | $1.6B | $10.7B |
| Datadog DDOGMonitoring, including of AI-native customers | $4.0B | +32% | 80% | 0% | $1.2B | $3.5B |
| HubSpot HUBSCRM and marketing software for mid-sized companies, with AI agents | $3.4B | +21% | 83% | 4% | $797M | $1.7B |
| Zscaler ZSCloud security that inspects traffic, including traffic to AI apps | $3.4B | +25% | 77% | −4% | $852M | $7.4B |
| Okta OKTAIdentity and access management, extending to AI agents | $3.1B | +11% | 78% | 8% | $979M | $4.9B |
| MongoDB MDBDatabase for application developers | $2.8B | +25% | 73% | −1% | $659M | $1.5B |
| Dynatrace DTObservability software with AI-driven analysis | $2.1B | +18% | 81% | 12% | $571M | $3.4B |
| Elastic ESTCSearch and vector database used in AI retrieval | $1.8B | +16% | 76% | −3% | $349M | $1.9B |
| Pegasystems PEGAWorkflow and decisioning software with generative AI | $1.7B | +1% | 76% | 10% | $442M | $2.0B |
| UiPath PATHAutomation software moving from robots to AI agents | $1.7B | +15% | 83% | 9% | $363M | $1.4B |
| Teradata TDCData analytics platform for large enterprises | $1.7B | +1% | 61% | 8% | $736M | $2.0B |
| CCC Intelligent Solutions CCCInsurance-claims software with AI photo estimating | $1.1B | +11% | 74% | 18% | — | $1.8B |
| GitLab GTLBDevSecOps platform with AI coding assistants | $1.1B | +23% | 86% | −9% | $214M | $1.2B |
| C3.ai AIEnterprise AI applications | $232M | −38% | 29% | −203% | −$156M | $200M |
| SoundHound AI SOUNVoice AI for restaurants, cars and customer service | $203M | +55% | 42% | −69% | −$116M | $60M |
Revenue and growth. Total revenue over the four quarters, and its change from the four quarters before. Fast growth with a flat margin usually means volume; fast growth with a rising margin usually means pricing power.
Gross margin. The share of revenue left after the direct cost of what was sold. It is the best single measure of how much of each dollar a company keeps rather than passes on to its own suppliers.
Operating margin. What is left after research, sales and overhead too. The gap between gross and operating margin shows how much the company spends to stay competitive.
Free cash flow. Operating cash flow minus capital spending: the cash the business generated after paying to maintain and grow itself. Negative free cash flow during fast growth means the growth is consuming cash.
Where the money comes from and goes
Its own spending flows to AI clouds. For the dollar amounts and who keeps them, see following the money.
What to watch in the next filings
- Revenue growth acceleration (or not) at the large established software companies as AI products launch.
- Changes in pricing models from seats to usage, and what they do to revenue per customer.
- Backlog (remaining performance obligations) growth as a sign of larger AI contracts.
- Gross margin at companies running heavy AI features: inference costs can compress it.
- AI labs’ revenue as disclosed through their partners’ filings.
Free · quarterly · from SEC filings
Get the AI Stack Report every quarter
After each Big Tech earnings season we update the map from the new filings and publish what changed: who spent, who got paid, who kept it and what the filings reveal about the loop. Free, in Rob May’s Investing in AI newsletter.
Form not showing? Subscribe on Substack. No spam; unsubscribe any time.
Questions about software
Are software companies making money from AI?
Some are. Growth across the group we track was +37% over two years, steady rather than accelerating. A few companies, notably Palantir and AppLovin, show AI clearly in faster growth and high margins; for many established vendors the reported numbers don’t yet show acceleration.
How does software revenue relate to AI capital spending?
Capital spending by the cloud companies is a cost that has to be repaid by revenue from AI users: cloud rental, subscriptions, advertising and software. Software companies are one of the places that revenue should show up if AI is making their products more valuable.
Could AI hurt software companies?
Yes. AI lowers the cost of building software, lets some tasks move to AI agents, and can shrink the number of seats customers pay for. Companies that price per user are most exposed unless they shift to pricing by usage or outcome.
Which software companies are growing fastest?
Over their latest four quarters, the fastest growers in the group we track were Palantir (+79%), AppLovin (+61%), SoundHound AI (+55%) and Shopify (+33%). Growth alone doesn’t show that AI is the cause; margins and backlog help tell.
How can I tell if a software company’s AI is working?
Look for it in the filings: faster revenue growth, rising margins, or a growing backlog of contracted revenue. Announcements and AI mentions on earnings calls are not evidence on their own. Our AI Signal Gauge is designed to check these from SEC filings.
Research and education, not investment advice: nothing here is a recommendation to buy or sell any security. Figures come from companies’ SEC filings as reported; companies named were chosen because they show how the stack works, not as a view on their shares.