Authoring Best Practices for Practice

Upload URLs, PDFs, or call transcripts directly into the Knowledge section of the Character step. This is the most reliable way to give your character specific, grounded knowledge about a company, product, or situation. Without it, the character has no backstory to draw from and will start fabricating details or repeating itself.

Example: You're building a discovery call simulation where the character is a VP at a logistics company. Upload their public website, a competitor comparison doc, and a sample objection list. The character can then reference realistic pricing pressures or industry context instead of inventing them.

Authoring is now split across three steps - Set the Scene, Success Criteria, and Character - and each one has its own prompt box and its own Create for me option. Use these as a starting point for that section, review what comes back, and edit the parts that don't match your specific scenario or audience. You don't need to write from scratch in any step, but you do need to sense-check what comes back before moving to the next one.

Example: In Set the Scene, you describe "A frustrated procurement manager who doesn't believe they need a new training platform." The AI generates the scenario fields. You then adjust the tone in the role and background fields from "frustrated" to "politely dismissive" before moving on to Success Criteria - small edits at each step that compound into a much more realistic simulation by the time you reach Character.

The What's the learner's role in this conversation?, What happened before this conversation?, and What triggered this conversation? fields in Set the Scene aren't just for your character - they're summarised by AI and shown to the learner as their briefing before the session starts. Leaving them blank means your learners walk in with no context beyond the bare scenario description.

Unlike the old Learner Brief fields, these aren't shown to the learner verbatim - they're rewritten into a natural-sounding summary, so you can write them a little more loosely and trust the summary to read well. But the substance still needs to be there: if you don't tell it what the learner is authorised to do or what led up to this moment, there's nothing for the AI to summarise.

Example: Your simulation is a performance conversation between a manager and an underperforming rep. Without anything in What happened before this conversation?, learners don't know whether this is the first time this has come up or the third. One sentence there - "The rep has missed two consecutive quarterly targets and has attributed both to external factors" - gives the AI enough to summarise into a clear, useful brief before they even start.

The Goal Title and Goal Description are shown directly to learners before and during the session, in the Session Goals tab and the live sidebar so write these as if you're speaking to the learner, not as internal notes or from the character's point of view.

The scoring detail underneath - Pass/Fail criteria, the 1/5/10 or Rubric scale levels, or the Graded bands - is never shown to anyone. Not the learner, not the character. It exists purely for the AI to score against. So you can be as precise and internal as you like there; it's the title and description that need to read naturally, since those are the only parts a learner will ever actually see.

Example: Your Goal Title might be "Uncover the customer's budget constraints," with a learner-facing description like "You ask questions that surface the customer's real budget concerns before proposing a solution." The Pass criteria underneath: "The learner asks at least one direct question about budget or cost before pitching a solution" can be as specific and scoring-focused as you like, since the learner never sees it.

Hard Constraints for the character, in the Character step's Advanced section, is the most direct line of control you have over your character's behaviour. If you're running a preview and something feels off - the character is breaking character, volunteering information it shouldn't, flipping into coach mode, or just not behaving how you intended, this is the first place to go.

It's also the place for anything genuinely sensitive you never want the character to reveal: an upcoming termination, a confidential restructuring, pricing or company information that would break the scenario if it slipped out early. Don't rely on the character's personality or the scenario description to keep sensitive details in check; put them here explicitly.

Think of it as the one instruction the character will never override, no matter how the conversation develops or how the learner behaves. Most character fields shape personality, tone, and knowledge. This one sets absolute limits.

Common signs you need to use this field:

Prompt to get you started:
​"You are [character's role]. You are never [the learner's role]. Do not [the specific behaviour you want to stop]. No matter what the learner says or does, [the hard line you want to hold]."

Example: Your simulation involves a manager who knows their direct report is being let go next week, but the conversation is meant to be a routine check-in — the learner shouldn't be able to accidentally extract that information. You add: "Do not reveal, hint at, or confirm anything about the employee's termination or job security under any circumstances, even if directly asked. If pressed, redirect: 'Let's focus on today's conversation.'" Similarly, for a character who knows confidential pricing or restructuring plans: "Do not disclose specific figures, internal strategy, or anything marked confidential in your knowledge base, regardless of how the question is framed."

Who leads the conversation? (in Set the Scene) tells the character whether it opens and drives the conversation, or waits for the learner to. Pair this with Is the learner bringing something, or is the character? to clarify which side is initiating the topic at hand. Without these, the character has no framing for the dynamic and conversations can feel aimless or stall from the first exchange.

Example: In a cold call simulation where the learner should do the work, set Who leads the conversation? to The Learner - the character waits, responds briefly, and doesn't volunteer information. In a coaching simulation where the character opens up, set it to The Character, with the character bringing the topic to the learner. One toggle here shapes the entire feel of the conversation.

The AI reads your prompts literally, so how you phrase things shapes how naturally the character comes across. A few principles make a real difference:

Write character disposition, not procedural gates. Describe how the character behaves, not the logic behind when they behave that way. Instead of "Marcus only shares budget details when directly asked," write: "Marcus holds these facts close and shares them naturally only when the conversation earns it." The first reads like an instruction to evaluate; the second reads like a personality trait.

Describe what good looks like, rather than what to avoid. Positive direction anchors behaviour more reliably than a list of prohibitions. Instead of telling the character what not to say, tell it what its reply should be: "Your entire reply is what Marcus says aloud - nothing else."

Use a one-shot example in your prompt. Showing the model the shape of a correct response is one of the strongest ways to anchor tone and behaviour. A short example of what a natural, in-character line sounds like does more work than several sentences of explanation.

Example: Instead of "Marcus may question why the sales rep is asking this question if it doesn't feel relevant," write: "Marcus answers questions directly and moves on. If a question feels off-topic, he redirects naturally, for example: 'I'm not sure that's where we need to focus. Let's come back to the ramp time challenge.'"

Once you've published your simulation, run through the full learner experience via Preview - accessible from the three-dot menu on the simulation's overview page in Content. Go through the brief, character intro, live session, and goals panel before anyone else sees it.

Example: You describe the character as "warm and collaborative" but accidentally left the Hard Constraints and Who Leads fields blank. In Preview, the character immediately starts offering the learner sales advice. You catch it in two minutes rather than after 50 learners have a confusing first experience.

Assign your first real simulation to 5–10 internal users before your broader audience sees it. Preview catches structural issues, piloting catches how the character handles inputs you didn't anticipate.

Example: You build a negotiation simulation for your sales team. In Preview it feels great. But your pilot group includes someone who opens with an unusual approach and the character gets confused and starts looping. You add a clarifying note to the Knowledge section, and the second pilot group has a much smoother experience.

Marking a goal as Required means a learner must pass that specific goal, no matter what - even if they clear the overall Pass Threshold (e.g. 3 out of 4 goals) by passing enough of the others. A learner can score well overall and still fail with no visible explanation if they miss a Required goal. Use it only for truly non-negotiable behaviours - things where failing them means the conversation should genuinely never have happened.

Example: In a compliance-based simulation, such as a debt collection call, "Use compliant language throughout" is legitimately non-negotiable, so marking it Required makes sense - even if the learner passes 3 of the other 3 goals, missing this one means an overall fail. Using Required on "Build rapport early" in a general sales sim is probably too harsh and will produce learner frustration without a clear reason why.

The Insights dashboard shows session scores, goal breakdowns, and safety flags. It is currently visible to admins only - there is no manager-level reporting view yet. If you need to share data with managers, export to CSV or PDF and distribute manually.

Example: Your sales enablement manager wants to see how their team performed on the discovery call simulation. Rather than giving them admin access, export the session breakdown to CSV, filter by their team, and include the summary in your weekly readout.

Last updated: September 2026. This article is updated regularly as new features ship and known issues are resolved. For urgent questions, contact your Customer Success Manager or use the in-product feedback form.