The Sword of Damocles in Software

Key Takeaways

• Copilot installs fell as Claude Code, Codex surged.
• Software NDR dropped from 125% to 112% (2022‑2025).
• 2026 bottom quartile NDR fell to breakeven levels.
• Simple SaaS products face AI-driven substitution risk.
• Growth needs >13% new acquisition to offset churn.

Pulse Analysis

The AI coding assistant market has become a bellwether for broader SaaS dynamics. When Claude Code and OpenAI Codex launched in mid‑2025, they captured developer attention with lower latency and tighter integration, prompting a swift shift in daily install patterns. This competitive shock illustrates how quickly a first‑mover advantage can erode when newer models deliver superior performance or cost structures, a lesson that extends beyond developer tools to any cloud‑based offering.

Beyond the coding niche, the data on net dollar retention reveals a systemic pressure on software firms. From 2022 to 2025, the industry‑wide NDR slipped gradually, but the 2026 inflection point—where the 25th percentile dropped to 101%—signals that the weakest quartile is now operating at break‑even. Companies whose products are easily replaceable—video conferencing, task management, SMB billing—are feeling the squeeze as AI agents replicate core functionalities at lower prices. This commoditization forces firms to either innovate with differentiated value or accept accelerated churn.

For executives, the implication is clear: growth strategies anchored solely on upselling existing customers are no longer sufficient. With a 96% NDR, Asana must acquire more than 13% new revenue annually just to maintain its 9% top‑line growth, a hurdle that intensifies as AI alternatives proliferate. Strategic responses may include investing in proprietary data, building AI‑enhanced features that are hard to copy, or pivoting toward higher‑margin, less replaceable verticals. Companies that fail to adapt risk joining the bottom quartile that is already bleeding revenue in a rapidly evolving AI‑driven landscape.