Skip to content

INNOVATION · SYNTHETIC CUSTOMERS

Use synthetic customers to challenge your B2B innovation adoption story

An innovation idea can sound useful while depending on a chain of changes that no customer has agreed to make. Someone must notice a problem, take ownership, change a routine and keep using the new approach. Which link is most uncertain?

For B2B innovation teams, synthetic customers are AI-generated customer perspectives that can help explore that question before a pilot. Virtual Customer supports structured conversations with AI-simulated B2B customers in a defined business context. Use those conversations to expose possible objections and missing dependencies, then decide what deserves direct customer research. The useful output is a sharper adoption story, not a simulated vote on whether the idea will succeed.

Illustrative virtual customer seated in a light jacket
Realistic virtual customersA business world. A role. A perspective.Explore in an ongoing text or voice dialogue.

Customer learning, throughout
the product journey

Explore needs

Test assumptions

Compare perspectives

Write the change you expect customers to make

Start with a behavior, rather than a product benefit. “Reduce downtime” is an outcome. “Review a new maintenance recommendation before scheduling the next shift” is a change in someone’s work.

Write one sentence containing the role, current routine, proposed change and intended outcome. Keep it visible throughout the exploration. If every conversation refers to a different version of the idea, apparent disagreement may come from the description rather than the customer context.

Also record what the customer would stop doing. An innovation can add work even when its promised outcome is attractive. The displaced spreadsheet, meeting or approval step may explain more than a general question about interest.

Use this chain as a working checklist for your team:

  1. Recognition: What makes the problem noticeable in an ordinary working week?
  2. Ownership: Who is expected to act, and who bears the consequence of doing nothing?
  3. Permission: What access, approval or cooperation is needed to try a different approach?
  4. First use: What must happen for the first useful result to appear?
  5. Repeat use: What would make the new routine worth continuing after the initial attention fades?

Attach an assumption to each link. “A supervisor will review the recommendation every morning” is more useful to investigate than “the experience will be intuitive.” It identifies a person, a trigger and an action that can later be observed.

Select the modeled customer’s geography, industry, company size and role. Begin with the person whose routine changes, then explore a second role that controls an important dependency. Keep the proposed innovation and operating context comparable.

Ask about current work before introducing your proposal. What triggers action today? What gets postponed? What information must be trusted? After explaining the idea, ask what would have to change for it to fit that routine.

Avoid asking the synthetic customer to endorse your conclusion. “What could make this impractical?” and “Whose cooperation would this require?” create more useful exploration than “Would this save you time?” Record alternative explanations instead of collecting favorable responses.

Example: a maintenance recommendation that nobody owns

Illustrative scenario, not a customer result.

An innovation team is exploring a maintenance-planning service for mid-sized manufacturers in Sweden. The service would suggest a maintenance window before equipment availability becomes a problem.

The team explores a maintenance manager’s perspective and a production supervisor’s perspective separately. The first conversation examines how recommendations become maintenance tasks. The second examines who can interrupt or rearrange production.

The critical question is the handoff: who turns a recommendation into an agreed change to the schedule? The team does not assume that either role can act alone. It prepares questions about the decision meeting, required information and circumstances in which a recommendation would be deferred.

No fictional response is treated as a finding. The example shows how to move from a broad benefit to a specific dependency worth investigating.

Turn objections into a research choice

Create a short adoption-chain note outside the tool. For each weak link, record the assumption, possible obstacle, consequence if wrong and the evidence that would resolve it.

Prioritize by how much the answer could change your next step. A wording concern may need a short comprehension check. An unresolved owner for a recurring task may require interviews across roles and observation of the existing handoff before a pilot is sensible.

This is a team judgment. Do not count repeated simulated agreement as independent support, or convert responses into an adoption percentage. Several fluent conversations can still share the same unsupported starting assumption.

Check the chain in real customer work

Use real conversations to establish whether the problem and responsibility exist in the intended context. Observe the relevant routine where possible. A bounded pilot can then test whether people actually complete the new action and obtain the intended result.

Synthetic customers help you prepare this work; they do not establish demand, operational feasibility or sustained adoption. Keep the original assumption beside the evidence so that a compelling explanation does not silently become a confirmed fact.

When one link changes, revisit its neighbors. A different owner can change permissions, first use and the reason to return. That makes the adoption chain useful beyond the first discussion.

Bring one B2B innovation idea, one customer context and one behavior that must change. Explore the dependency you understand least, then choose the real-world check that could change your pilot decision.