The three questions that decide your CPQ vendor

Question 1: what does your product model actually look like?

Some products are rule‑heavy, thousands of "if A then not B" statements between options. Cranes, elevators, industrial machinery. These map naturally to constraint‑solver CPQ platforms like Tacton or Configit, where the model is a network of relationships and the solver figures out what is valid.

Other products are flow‑heavy, the sale is a sequence of steps (discovery, sizing, pricing, approvals) with only mild option interdependence. Subscription hardware, medical devices, telecom bundles. These map naturally to workflow CPQs like Salesforce Revenue Cloud or Oracle CPQ, where the strength is quote process, not constraint solving.

Trying to model rule‑heavy products in a workflow CPQ is where most bad selections happen. The demo looks fine. Month nine looks catastrophic.

Question 2: what does your integration reality look like?

Every CPQ vendor has a "certified" connector to every major ERP. Almost none of them work out of the box in production. What matters is which vendor's partner ecosystem has done your specific integration recently.

If you are on SAP S/4HANA with a heavily customised variant configuration, your realistic shortlist is different from a company on Microsoft Dynamics 365 with vanilla item master data. Ask each vendor for three references in the last 18 months on your exact ERP version and integration pattern. The list gets short fast.

Question 3: what modelling capacity can you commit?

The uncomfortable question. A CPQ model is code. Someone has to write it, maintain it, and evolve it as products change. Vendors will tell you their tool is "business‑user friendly" and does not need a developer. This is partially true and dangerously misleading.

A working truth: you need at least one internal person who owns the model end‑to‑end, and either a partner or a second internal person for capacity. If you cannot commit that, no vendor choice will save the programme. If you can, most reasonable vendors will succeed.

What RFP scorecards miss

Scorecards weight features. Features are the easiest thing to demo and the least predictive of outcome. The three questions above weight structure, context and capacity, the things that actually determine whether the model is still working three years after go‑live.

If you want a second opinion on a shortlist you are already down to, that is exactly what we do. Get in touch.