So why are we talking so much about Cursor, Claude Code, etc?

Well for a good reason. Not only did both rocket to $1 B ARR in less than one year, no it’s more than that.

Coding is 55 % of all departmental AI spend, per the latest Menlo Ventures analysis.

Image: Departmental AI spending by category, dominated by product and engineering with $4.2 billion, followed by IT at $700 million, marketing at $660 million, customer success at $630 million, sales at $390 million, HR at $360 million, data science at $240 million, finance and operations at $100 million, and market research at less than $100 million.

55 % of All Departmental AI Spend Is Now on Coding. And It’s Not Slowing Down.

That’s $4.0 billion out of $7.3 billion total.

Not AI SDRs or chatbots. Not content generation. Not customer support. Coding.

This is why Cursor, Claude Code, Replit, Supabase and the whole AI coding ecosystem are exploding. It’s not hype. It’s where the money is actually going.

The Scale of What’s Happening

Departmental AI spending hit $7.3 billion in 2025, up 4.1× year‑over‑year. Coding alone went from $550 M in 2024 to $4.0 B in 2025—a 7× growth in twelve months.

• 50 % of developers now use AI coding tools daily; at top‑quartile organizations it’s 65 %.

• Code completion hit $2.3 billion.

• Code agents and AI app builders exploded from near‑zero.

• Teams are reporting 15 %+ velocity gains across the software development lifecycle—from prototyping to refactoring to QA to deployment.

• Cursor hit $200 M in revenue before hiring a single enterprise sales rep. Pure PLG. Developers found it, loved it, and pulled it into their companies.

Image: AI coding spend is projected to increase significantly from 2024 to 2025, across multiple areas.

But What About Everything Else?

55 % going to coding doesn’t mean the other categories don’t matter. It means coding figured it out first.

IT Operations — 10 % (~$730 M)

Automating incident response, infrastructure management, monitoring. This is the natural second act for AI in technical organizations. If AI can write the code, it can help you keep it running. Still early, but the use cases are obvious and the budgets are there.

Marketing — 9 % (~$660 M)

Content generation, campaign optimization, personalization at scale. This is where most non‑technical teams first touched AI—writing emails, generating ad copy, creating variations. The challenge? Harder to measure ROI than “did the code ship faster.” But the spend is real and growing.

Customer Support and Success — 9 % (~$630 M)

Ticket routing, sentiment analysis, proactive outreach, knowledge‑base automation. Support teams are adopting fast because the metrics are clear: time‑to‑resolution, tickets per agent, customer‑satisfaction scores. This category has the clearest path to a “Cursor moment”—measurable, high‑volume, workflow‑native.

Design — 7 % (~$510 M)

Still finding its footing. AI design tools exist, but they haven’t cracked the workflow the way coding tools have. Figma with AI features is one thing. A design tool that makes designers 15 %+ faster the way Cursor makes developers faster? Not there yet. Watch this space.

HR — 5 % (~$365 M)

Recruiting automation, employee engagement, policy Q&A, onboarding. Smaller category, but steady. HR tech has always been a slow adopter, and AI is no exception. The wins here are real but incremental.

Other — 5 % (~$365 M)

Everything else—legal, finance operations, procurement, facilities. These functions are buying AI tools, but spend is fragmented across dozens of niche solutions.

Why Coding Won First

The Menlo report points to a specific inflection point: Anthropic’s Claude Sonnet 3.5 in mid‑2024. That’s when AI coding tools became “economically meaningful”—good enough that the productivity gains justified the spend at enterprise scale.

But there’s a deeper reason coding won first: it’s measurable.

• Lines of code shipped.

• Pull requests merged.

• Time from ticket to deployment.

• Bugs caught before production.

These are numbers engineering teams already track. When AI coding tools moved those numbers, the business case wrote itself.

The categories that will break out next are the ones that can prove ROI just as clearly. Customer Success has ticket‑resolution times. Marketing has campaign performance. HR has time‑to‑hire.

The categories that will struggle are the ones where “AI made it better” is hard to quantify.

The Bottom Line

• 55 % of departmental AI spend = $4 billion, 7× year‑over‑year growth.

• Coding is generative AI’s first killer use case. That’s why the Cursors, Claude Codes and Replits of the world are exploding.

• The other categories aren’t irrelevant—they’re just earlier. The $730 M in IT operations, the $660 M in marketing, the $630 M in customer success—that’s real money, and it’s growing fast.

• If you want to understand where enterprise AI is today, follow the developers. They figured it out first.

---

Source: Menlo Ventures 2025 State of Generative AI in the Enterprise