Clean CPQ data is becoming the silent precondition for agentic selling, and the gap between companies with it and companies without it is about to widen.
An analysis published this week by SalesTechStar examines why AI sales agents fail in complex selling environments: not because of model limitations but because of bad product logic in the underlying systems. When CPQ configurations are inconsistent, pricing rules are underdocumented, or product constraints are missing, AI agents surface invalid configurations, generate error-prone proposals, and require manual intervention to fix. The automation fails precisely at the point where it would create the most commercial value.
Threekit’s AI Sales Agent launch this week, aimed at manufacturers selling complex configured products, is a partial response to this dynamic. The agent accepts inputs from voice memos, photos, RFPs, and engineering drawings, generating valid configurations without requiring complete specifications. It reported 229 percent revenue growth in the past year and handles more than 10 million configuration-to-quote sessions monthly. But the agent’s accuracy depends on the underlying product data being structured correctly to begin with.
For sales leaders evaluating agentic investments in complex-sale environments, the CPQ layer is the prerequisite audit. Companies that have already modernized their product configuration and pricing logic, as documented in the recent wave of CPQ modernization, are better positioned for agentic deployment. Those that have not will find that agents surface the data quality problem faster than they solve it.
The original insight: infrastructure debt in CPQ is now a deployment gate for AI, not just a technical liability. The revenue teams with the cleanest product data will see the fastest agentic ROI. Everyone else has a sequencing problem.
Source: SalesTech Star
