How to use Glyphic for rep coaching

Scorecards automatically evaluate every call against your team's own standards, surface coaching opportunities, and track skill development over time. This guide walks through how to design, deploy, and get the most out of scorecards, whether you're running a sales team, a CS org, or a compliance-sensitive operation.

How scorecards work

The system has three layers:

1. Call types (Tags) — categories like "Discovery Call", "Demo", "Enterprise Kickoff", or "Cold Call". Glyphic automatically tags each call based on its content. You write a short definition for each type so the AI knows what to look for.

2. Scorecards — a scorecard is attached to one or more call types. It contains the skills you want to evaluate for that type of conversation. Different call types get different scorecards because expectations vary.

3. Skills with rubrics — each skill has a name, a written definition, and a scoring rubric (typically 1–5) describing what each level looks like. Glyphic reads the transcript and scores each skill against your rubric.

When a call is processed, Glyphic identifies the call type, finds the matching scorecard, and scores every skill automatically. The overall score is a weighted composite of all skill scores.

Step 1: Define your call types first

Before building scorecards, set up your call types using Tags. Getting the auto-tagging right is foundational — everything else depends on it.

Go to Settings > Tags and create a tag for each distinct type of conversation your team has. Write a clear definition so Glyphic can classify calls accurately. Common setups include:

Tip: Start with 3–5 call types that cover 80% of your conversations. You can always add more later.

If the auto-tag gets a call wrong, anyone can manually override it from the call page. The call will automatically rescore against the correct scorecard within minutes.

Step 2: Design your scorecards

Navigate to Coaching > Scorecards and create a scorecard for each call type. For each scorecard you'll define:

• Name — e.g. "Discovery Rubric" or "Enterprise Kickoff Scorecard"

• Call tags filter — which call types this scorecard should evaluate

• Skills — the individual competencies you want to measure

• User/team filter (optional) — restrict to specific reps or teams

Designing skills and rubrics

Each skill needs a definition (what does this skill mean?) and a rubric (what does each score level look like?). The rubric is a series of descriptions — one per level — that tell the AI exactly what to listen for.

A typical 5-level rubric moves from "did not attempt" through to "exceptional execution". For example, a Discovery Call skill called Pain Point Discovery might have:

• 1 — Rep did not ask about challenges or pain points

• 2 — Surface-level question asked but not followed up on

• 3 — Rep identified one clear pain point and explored it briefly

• 4 — Multiple pain points uncovered with follow-up questions

• 5 — Deep, multi-layered discovery uncovering second and third-level pain points with quantified business impact

Tip: If you're not sure where to start, use the AI-generate option. When creating a scorecard, you can ask Glyphic to generate skills and rubrics based on a description of the call type. You can then edit the output to match your methodology.

Glyphic also ships with default rubric templates for common scenarios:

• Discovery skills — Pain Point Discovery, Value Proposition Alignment, Qualification, Rapport, Objection Handling, Active Listening, Next Steps

• Demo skills — Recap, Product Demo, Customization, Handling Technical Questions, Closing

• Negotiation skills — Negotiation Approach, Value Articulation, Flexibility, Persistence

Weighting skills

Not every skill is equally important. Each skill has a multiplier (default 1.0) that controls how much it contributes to the overall score. If "Setting Clear Next Steps" is twice as important as "Rapport" for your team, give it a multiplier of 2.0.

Separating compliance from coaching

If your team has compliance requirements (e.g. mandatory disclosures, consent language), create a separate universal scorecard that applies to all calls — don't mix compliance criteria into your coaching scorecards. This keeps coaching focused on skill development while ensuring compliance is tracked independently.

For compliance items, consider using a 2-level (pass/fail) rubric rather than a 5-level gradient, since compliance is typically binary.

Screenshot needed: The "Create Scorecard" or "Edit Scorecard" page showing the skill editor with a rubric being defined — score levels 1 through 5 with descriptions, plus the multiplier field.

Step 3: Roll out in phases

We recommend a phased approach based on what we've seen work best across teams:

1. Week 1–2: Managers only. Enable scorecards for your managers first. Let them review scores against their own knowledge of calls to validate that the AI's scoring aligns with their judgment. Tweak rubrics and definitions where needed.

2. Week 2–3: Refine. Adjust call type definitions if tagging accuracy is off. Tune rubric descriptions — the more specific you are about what "great" looks like, the more accurate the scoring. This calibration period is normal and expected.

3. Week 3+: Reps. Once managers are confident in the scoring, roll out visibility to reps. Set expectations that scores are a development tool, not a punishment mechanism.

Important note: The AI is calibrated to find areas for improvement. Scoring 100% is very rare and genuinely impressive. Set that expectation early to avoid discouragement.

Using the coaching dashboard

The Coaching section gives you three views:

Scorecard overview

See all your scorecards with aggregate scores. Click into any scorecard to see individual skill performance across the team.

Team view

A bird's-eye view of all reps across all scorecards. Filter by team, scorecard, or date range. You can:

• See trend lines showing whether skills are improving or declining

• Compare individual rep averages against the team average

• Hover over data points to see score explanations

• Filter by quarter for performance reviews

Screenshot needed: The Coaching team view showing a line chart of skill scores over time for multiple reps, with the team average line visible. The scorecard and date range filters should be visible at the top.

Individual rep view

Click into any rep to see their personal coaching dashboard:

• Radar chart — visual snapshot of strengths and focus areas across all skills

• Score history — trends for each skill over time

• Super powers — skills where the rep consistently scores highly (55%+ threshold)

• Focus areas — skills that need the most improvement (below 55%)

• Call-level drill-down — click any score to see the full explanation, improvement suggestions, and jump to the exact moment in the recording

Screenshot needed: An individual rep's coaching page showing the radar chart on the left and the skills breakdown on the right, with "Super Powers" and "Focus Areas" sections visible.

How reps use scorecards day-to-day

On every call page, reps see their scorecard results alongside the call summary. For each skill:

• Hover over the score to see why they received that rating

• Scores of 3 or lower include specific improvement suggestions

• Click the play button to jump to the exact moment in the recording where the skill was demonstrated

• Use "Watch an example" to see a colleague's call where they excelled at that specific skill — the recording jumps straight to the relevant timestamp

This last feature is especially powerful: it creates a self-serve coaching library curated by the AI from your team's real calls.

Automated coaching summaries

Managers can receive a weekly coaching summary via email or Slack (typically sent on Fridays). The summary includes:

• Top skills across the team

• Focus areas that need attention

• 2–3 reps who either overperformed or had an off week

• Trends and shifts compared to the previous period

To enable summaries, go to your notification preferences and toggle on coaching summary delivery for email, Slack, or both.

Integrating scorecards with your sales methodology

Scorecards work best when they reflect the methodology your team already follows. Here's how to align them:

Using Agents for coaching insights

The Agents feature can pull coaching data to answer questions that go beyond what the dashboard shows. Useful prompts include:

• "How has [rep name]'s discovery scoring trended over the last quarter? What are their biggest areas of improvement?"

• "Compare coaching scores between Team A and Team B for demo calls this month."

• "Which reps have the lowest 'Setting Next Steps' scores, and what specific feedback are they receiving?"

• "Generate a coaching report for my team's performance last quarter that I can share in our all-hands."

Agents have access to all scorecard data, so you can ask natural-language questions about performance trends, skill gaps, and team comparisons without needing to export or manipulate data manually.

AI Training Simulator

For teams that want to go further, Glyphic's AI Training Simulator lets reps practice specific call types before going live. The simulator:

• Creates realistic AI personas based on your actual customer profiles and call data

• Lets reps run through discovery, kickoff, or demo scenarios

• Scores the practice session using the same scorecards applied to real calls

• Reports practice scores to managers, making it easy to check comprehension without one-on-one time

This is particularly valuable for new hire onboarding — establish a baseline scorecard performance during ramp, then track how quickly new reps converge with the team average.

Visibility and permissions

By default, reps can see their own scores. Managers can see scores for their direct reports. Admins can grant broader visibility using the View Others' Scorecards permission if needed.

Even with restricted visibility, managers can send individual call snippets to reps to highlight strong examples or coaching moments.

Exporting scorecard data

Coaching data can be:

• Exported to Excel — useful for quarterly performance reviews

• Synced to your CRM (HubSpot or Salesforce) — coaching scores appear alongside deal and contact records for integrated reporting

Common mistakes to avoid

• Too many skills per scorecard. Start with 5–7 skills per call type. More than 10 makes it hard for reps to focus.

• Vague rubric descriptions. "Good discovery" is not specific enough. Describe exactly what the rep should say or do at each level.

• Mixing compliance and coaching. Keep compliance in a universal scorecard. Coaching should be about skill development.

• Launching to reps without validation. Always have managers review scoring accuracy for 2–3 weeks before making scores visible to reps.

• Treating scores as punishment. Frame scorecards as a development tool. Highlight improvements and celebrate high scores, not just flag low ones.

Need help?

If you'd like help designing scorecards tailored to your methodology or reviewing your rubrics, reach out at support@glyphic.ai. We regularly help teams get their scorecards right during onboarding.