How Skillet turns a roleplay conversation into a score: what your accuracy score reflects, how competencies and objections are evaluated, what is measured but kept separate (soft skills), and why you can trust the results.
In this article
Your accuracy score, at a glance
The headline result of every roleplay is your Accuracy score, shown as a percentage from 0 to 100. It reflects how closely your conversation matched what the scenario was designed to assess: the competencies it evaluates, the objections it expects you to handle, and the key points it wants you to cover. The higher your accuracy, the closer you came to handling that scenario well.
Your score appears with a short performance summary, and on a reattempt it shows the change from your last attempt.
Competencies and sub-competencies
Skillet evaluates your conversation against competencies: the selling skills that matter for the scenario, such as exploring their needs, objection handling, clinical evidence mastery, disease knowledge, product knowledge, and gaining commitment.
Each competency breaks down into finer sub-competencies, so your feedback is specific rather than general. The scenario decides which competencies apply, so you are assessed on what is relevant to that conversation, not a fixed checklist.
Your report’s competency scorecard shows how you did on each competency and sub-competency, with AI feedback and “areas to explore” (see Reading your roleplay report).
How objection handling is scored
When your HCP raises a concern, Skillet tracks how completely you handle it through four stages:
Raised: the HCP brings up a concern.
Clarified: you make sure you understand what is driving it.
Addressed: you respond with relevant, approved information.
Closed: you resolve the concern and confirm the HCP is satisfied.
Your report shows each objection, how far you took it, and how many you fully closed (for example, “2 / 2 objections successfully closed”).
Covering the key points
A scenario can define key points you are expected to cover, such as specific approved messages or clinical claims. Your evaluation reflects whether you covered them in the conversation.
Soft skills: measured, but separate from accuracy
Skillet also measures your soft skills: how you deliver the conversation, not what you say. These are reported separately and are not counted toward your accuracy score, so delivery style never inflates or deflates your result.
The delivery measures Skillet looks at include your pace, engagement, how long you speak before pausing (monologue length), your talk-to-listen ratio, filler words, interruptions, and pauses.
Keeping soft skills separate means your accuracy score reflects message fidelity (whether you delivered the right, approved content effectively), while you still get actionable feedback on how you came across.
Meeting the agenda
Every scenario has an agenda and a goal. Your report’s performance summary shows whether you met the agenda (for example, “Agenda met,” “Agenda partially met,” or “Agenda not met”), along with the outcome of the call.
How the AI evaluates your conversation
Skillet evaluates the full transcript of your conversation against the scenario’s defined criteria: its competencies, the objections it expects, the key points to cover, and its agenda.
Crucially, the AI works within a closed knowledge ecosystem: it judges your conversation against your organization’s approved, loaded content, not the open internet. That is what keeps feedback consistent and on-message: two reps practicing the same scenario are held to the same standard. (For how Skillet handles and protects your data, see Security, Privacy & Compliance.)
Seeing this in your report
All of this comes together in your roleplay report. To learn how to read each part (the scorecard, objections, soft skills, transcript, and recording), see Reading your roleplay report.
Related articles
Reading your roleplay report
Run the roleplay: setup, the live call, and what happens if it ends early
The Analytics dashboard: trends, performers, and team pivots
Glossary of Skillet and life-sciences terms
Was this article helpful?
That’s Great!
Thank you for your feedback
Sorry! We couldn't be helpful
Thank you for your feedback
Feedback sent
We appreciate your effort and will try to fix the article