Sales managers and learning and development leads who run one-size-fits-all coaching programs see the same result: strong analytical learners disengage from roleplay, kinesthetic reps ignore written feedback, and the coaching dashboard shows completion rates that say nothing about behavioral change. This guide shows you how to use call data to identify each rep's learning profile and configure AI coaching assignments that actually match how they learn.
What is the 70/20/10 rule in coaching?
The 70/20/10 model holds that 70% of professional development comes from on-the-job experience, 20% from peer interaction and feedback, and 10% from formal instruction. For sales and contact center coaching, this means the bulk of learning should happen through call review and practice (the 70%), with AI-assisted roleplay and peer discussion supporting the 20%, and structured training content filling the 10%. A coaching program weighted toward formal training modules at the expense of call-based practice inverts the model.
Step 1: Audit Your Call Data for Learning Style Signals
Learning style is not a self-report exercise. Reps rarely know how they learn best, and asking them produces socially acceptable answers rather than accurate ones. Call recordings reveal learning gaps behaviorally.
Pull a 30-day sample of calls per rep and look for four patterns:
- Analytical learners miss criteria that require interpretation (empathy, urgency calibration) but score well on structured compliance items. They follow process but miss nuance.
- Auditory learners perform well when they have heard a model call recently but drift when they have not had recent exposure to examples.
- Kinesthetic learners show high score variance. They improve sharply after live practice sessions but regress without continued reps.
- Visual learners improve after seeing their own scored transcript side-by-side with a top performer's. Abstract feedback does not move their behavior.
Insight7's QA scoring platform generates per-criterion, per-rep scorecards across 100% of calls, giving you the behavioral pattern data needed to make this segmentation. Manual QA at 3 to 10% call coverage cannot produce a reliable learning profile because the sample is too small to distinguish style from random variance.
Avoid this common mistake: grouping reps by tenure rather than behavioral pattern. A five-year rep can be a kinesthetic learner who has coasted on muscle memory and regresses without regular practice just as easily as a new hire.
Step 2: Map Learning Gaps to the Right Coaching Format
Once you have identified learning style signals in the call data, map each gap to the format most likely to close it.
| Learning Profile | Gap Signal in Calls | Recommended Format |
|---|---|---|
| Analytical | Low empathy / tone scores despite process compliance | Transcript annotation + criterion-level explanation |
| Auditory | Inconsistent performance without recent model call exposure | Model call libraries + recorded feedback messages |
| Kinesthetic | High variance; drops after coaching gaps | Frequent short roleplay sessions; unlimited retakes |
| Visual | No improvement from verbal feedback alone | Side-by-side scored transcript comparisons |
Insight7 supports each of these formats. Roleplay sessions run on web and mobile (iOS), with voice-based post-session reflection that engages reps in discussion rather than just delivering a scorecard. Transcript evidence is embedded in every QA score so managers can annotate specific moments and explain the gap in writing for analytical learners.
Step 3: Configure Coaching Assignments Per Learning Profile
With gap data and format mapping in hand, build the actual assignments. The configuration decisions that matter most are scenario source, session length, and retake policy.
For kinesthetic learners: generate practice scenarios directly from real calls using Insight7's AI coaching tools. The hardest objection closes in your call library become the practice scenarios. Set retake thresholds rather than a single pass/fail. Reps see their improvement trajectory (40 to 50 to 80) as they practice, which is itself motivating for this profile.
For analytical learners: weight written feedback heavily. In Insight7, the criteria context column defines what "good" and "poor" look like for each item. Sharing that context with analytical learners directly bridges the gap between their process-following strength and their nuance weakness.
For auditory learners: build a curated model call library. Tag calls by scenario type so reps can self-select before handling similar situations. Supplement with voice-based post-session coaching that sounds like a conversation, not a form.
For visual learners: use side-by-side transcript comparisons between the rep's scored call and a top performer's call on the same criteria. Seeing their own language next to effective language is more instructive than any abstract feedback message.
What are the 5 C's of coaching?
The 5 C's framework covers: Clarity (what behavior to change), Context (why it matters), Criteria (what good looks like), Commitment (the rep's agreement to practice), and Check-in (follow-up measurement). AI coaching platforms operationalize all five when configured correctly. Criteria and clarity come from the QA scorecard; context comes from transcript evidence; commitment comes from assignment acknowledgment; check-in comes from score tracking over time.
Step 4: Build Practice Scenarios From Real Calls
Generic roleplay scenarios have a short shelf life. Reps quickly recognize the script and optimize for the scoring rubric rather than the real skill. The most durable scenarios come from actual call recordings.
Insight7 lets you generate roleplay scenarios from transcripts directly, turning the hardest real-world closes into objection-handling practice. The AI persona on the other side of the practice call is configurable by communication style, emotional tone, empathy level, and assertiveness, so you can match the customer type a specific rep struggles with rather than running every rep through the same generic scenario.
For kinesthetic learners, set no ceiling on retakes and track the improvement trajectory. Reps retaking sessions until they pass a configured threshold show measurable skill building, not just compliance.
For analytical learners, add a manual scenario configuration step where the manager or L&D lead annotates what the scenario is specifically testing. That framing helps analytical reps engage with the nuance rather than defaulting to their process-following mode.
Step 5: Deploy at Scale With Manager Visibility
Personalized coaching at the individual level is only sustainable at scale if managers have a single view of team progress. Without aggregated visibility, personalization creates administrative overhead that kills adoption.
Insight7's coaching dashboard supports bulk assignment to entire teams from a single interface while preserving individual scenario configuration. Supervisors approve AI-suggested training before it deploys, keeping a human in the loop without requiring manual scenario creation for every rep.
Track four things at the manager level: assignment completion rate by rep, score trajectory across retakes, which criteria improved after each coaching cycle, and which learning profiles are responding to their assigned format. If kinesthetic learners are completing sessions but not improving, the scenarios need to be harder or more varied. If analytical learners are not completing, the format is probably too heavy on audio feedback.
Step 6: Iterate Based on Score Trajectory and Behavioral Change in Calls
Coaching completion is a leading indicator. The lagging indicator is QA score change on the criteria that were targeted. Close the loop monthly.
Pull QA scores for the criteria you assigned coaching against. Compare the 30-day post-coaching score to the 30-day pre-coaching baseline. Separate results by learning profile to see which format-gap combinations are working.
Insight7's platform tracks rep score improvement over time, connecting the QA trend line to coaching history. That connection is what transforms coaching from a morale activity into a measurable performance program. When a board-level question comes up about training ROI, the answer should be: agents who completed profile-matched coaching improved their target criteria scores by X points over 60 days.
Revise assignments quarterly. Learning needs shift as reps gain experience and as your product or process changes. An empathy gap in Q1 may become a compliance gap in Q3 after a policy update. The call data will tell you what changed.
FAQ
Can AI coaching really adapt to individual learning styles automatically?
Not fully automatically. The AI surfaces the behavioral patterns in call data and generates practice scenarios; the manager or L&D lead still interprets the learning profile and selects the appropriate format. The configuration decisions described in this guide require human judgment. What AI does is give you 100% call coverage and consistent scoring so that judgment is based on real behavioral data rather than impression or anecdote.
How many coaching sessions per learning profile per month is optimal?
Research on skill retention suggests that short, frequent practice sessions outperform long infrequent ones regardless of learning style. For kinesthetic learners, two to three short roleplay sessions per week is more effective than one long session per month. For analytical learners, one annotated transcript review per week with a follow-up reflection discussion is often sufficient. Audit your data quarterly to see what cadence correlates with score improvement on your team specifically.
What if a rep does not fit neatly into one learning profile?
Most reps show a dominant pattern with secondary traits. If a rep's call data shows both high analytical compliance and high kinesthetic variance, start with the format that addresses the larger point gap. After the primary gap closes, reassess. Mixed profiles often resolve into a clearer dominant type once the most obvious behavioral gap is addressed through targeted practice.
