In-person roleplay: practice for meetings that happen face to face

A rep can run a flawless discovery call from a headset and still go quiet at the counter when a customer walks up mid-sentence and asks about price. Phone and video roleplay never rehearsed that moment, because it isn't a phone or video moment at all.

‍

Why this matters

A large share of sales and service conversations never happen over a call. They happen on a retail floor, at a property during a showing, across a desk in an insurance branch, or on-site with a field customer. Those settings run on different rules: a rep has to read posture and hesitation, manage silence without a "yes" or "mm-hmm" to fill it, and handle interruptions a phone conversation never produces. Training built around voice or video calls doesn't prepare anyone for that, and until now, Outdoo's roleplay agents assumed a call was the default context.

‍

How it works

‍

‍

In Person is a call type, not a separate product. When creating or editing a roleplay agent, an admin sets Call Type to In Person. The session still runs as a live voice conversation with the AI buyer, but the scenario framing, rep instructions, and scorecard shift to match a physical setting, accounting for pauses and room-reading rather than the tighter back-and-forth of a phone call.

‍

Multi-phase scenarios carry over. If a rep has to get past a gatekeeper, such as a receptionist, before reaching the actual meeting, that phase and the in-person phase are scored on two separate scorecards, so "got past reception" and "handled the meeting well" stay independent. Background noise settings apply to the gatekeeper portion, keeping the in-person phase focused on the face-to-face exchange itself.

‍

A manager who sat in on the actual meeting can also add a manual score to the session afterward, the same way they can for a voice call, so a coach's own read on the room can sit alongside, or override, the AI-generated score. Full setup steps are in Intro to in-person roleplay agents.

‍

What this unlocks

A retail team can rehearse the exact walk-in moment floor reps face daily: a browsing customer who hasn't asked for help yet, an opening that has to land without a script prompt to lean on.

‍

A real estate team can practice a site visit where questions about the neighborhood and closing timeline arrive in whatever order a real buyer actually asks them.

‍

An insurance team can run in-branch needs analysis and plan comparisons across every branch under an identical scorecard, so a regional manager can compare readiness location to location. Any field team meeting customers on-site, from account visits to service calls, gets a rehearsal environment shaped like the room they'll actually be standing in.

‍

A trade show team can rehearse the booth conversations that don't look like any of the above: a prospect who stops for thirty seconds between sessions, expects an elevator-length pitch, and walks off if the rep can't land the value before the next demo starts. That's a different rehearsal problem than a scheduled meeting, and it's one voice or video call practice never modeled either.

‍

The bigger picture

Roleplay software defaulted to the call because the call was the easiest thing to simulate: two voices, a script, a timer. But retail, real estate, insurance branches, and trade show floors were never call-shaped, and treating them as calls anyway is why so much face-to-face training still happens by shadowing someone rather than by rehearsing. Most training still treats the call as the unit of practice.

‍

This treats the room as the unit instead, which is where a surprising share of deals and trust actually get decided.

‍