AI-native software grows faster, but embedded AI has stronger economics

AI-native software companies reported faster growth than peers in 2025, but companies embedding AI into existing products posted stronger profitability and revenue productivity, according to KeyBanc Capital Markets' 17th annual benchmark. The findings suggest that AI adoption is broad, while monetization and measurable financial impact remain early.
- AI-native companies reported 34% median ARR growth in 2025.
- Embedded AI companies reported a median EBITDA margin of (1%).
- AI-driven ARR is projected to grow 128% in 2026.
- 44% of respondents said AI spending was over budget year to date.
- Only 20% of respondents offer fully autonomous agents to customers.
AI-native companies lead on growth
KeyBanc Capital Markets, the corporate and investment banking unit of KeyCorp, expanded its former SaaS Survey this year to examine how artificial intelligence is changing product strategy, monetization, unit economics and operating models in private software companies.
Companies selling AI as their primary product reported median annual recurring revenue growth of 34% in 2025. That compared with 22% for companies embedding AI in broader products and 9% for companies still developing customer-facing AI.
The growth lead came with weaker financial performance. Core AI product companies reported a median EBITDA margin of (12%) in 2025, compared with (1%) for embedded AI companies. Embedded AI companies also reported $201K in ARR per full-time equivalent employee, versus $117K for core AI product companies.
Companies charging separately for AI products reported 27% median ARR growth in 2025, compared with 17% for companies that did not. KeyBanc said that advantage had not yet translated into higher net retention.
AI revenue is rising as costs pressure margins
AI-driven ARR is projected to grow 128% in 2026, up from 74% in 2025, but it is still expected to represent only 12% of total ARR. Across all companies surveyed, median ARR growth was 24% in 2025 and is projected at 23% in 2026. Participants are planning for broad-based growth of 40% in 2027.
Median EBITDA margin across the benchmark is expected to improve from (3%) in 2025 to 1% in 2026. At the same time, 44% of respondents said AI spending was over budget year to date, while 24% said spending was more than 20% over budget. Higher-than-expected token usage was the most frequently cited reason for the variance.
Companies spending more than 5% of revenue on AI reported a median gross margin of 66% in 2025, compared with 78% among those spending less than 2%.
“AI is accelerating growth and compressing development cycles, but AI spend is running ahead of budget and the economics are still catching up,” Scott Peterson, managing director of Software Investment Banking at KeyBanc Capital Markets, said in the announcement.
Development speeds up, while workforce effects remain ahead
The benchmark found that 92% of respondents said AI was shortening product development cycles, and 91% said they deploy AI internally in engineering. KeyBanc said 65% of participants reported no AI-related headcount reduction or reallocation in 2025.
That picture is expected to change over the next 12 to 24 months. The share reporting no reduction or reallocation falls to 34%, with research and development at 45% and customer service at 37% identified as the functions most likely to be affected.
Customer use is also influencing software strategy. Some 57% of respondents said their customers were building alternative products with AI coding tools, while only 20% offered fully autonomous agents to customers.
Peterson said developer productivity gains were being reinvested into shipping more product faster, rather than flowing directly to profits. For software users and investors, the benchmark frames 2026 as a test of whether rapid AI-led growth can develop into durable revenue and stronger economics.


