Using Call Transcripts to Improve Coaching Calls
Sales managers and contact center team leads who run coaching sessions from memory are working with a structural disadvantage. "I listened to a few calls and noticed you do X" is an impression, not evidence. Transcript-based coaching replaces that impression with a specific quote, a timestamped moment, and a criterion-level score. The agent can no longer dispute the observation, and the manager no longer needs to defend a feeling. According to ICMI research on contact center coaching, agents who receive specific, behavior-level feedback tied to documented call moments improve targeted skills at significantly higher rates than agents who receive general performance summaries. Are there call coaching bots available for transcript analysis? Yes. AI-powered coaching platforms like Insight7 analyze call transcripts automatically and generate scored coaching feedback without requiring a manager to manually review each call. These systems go beyond summarization to evaluate specific behaviors against a coaching rubric, flag patterns across multiple calls, and route targeted practice scenarios to reps. The difference from a basic transcription bot is that the analysis is structured against your team's specific criteria rather than producing generic summaries. What you need before the first session Before running transcript-based coaching, you need scored call recordings from the past two to four weeks, at least three to five calls per agent, a scoring rubric with named criteria (not just a total score), and the ability to pull the specific transcript quotes that triggered each score. Set aside 30 minutes of preparation time per agent, which is what makes sessions more efficient rather than longer. Step 1: Pull 3 to 5 Scored Calls and Identify 2 to 3 Transcript Moments Per Call Select calls from the past two to four weeks that are already scored. Choose calls containing clear examples of the specific behavior you plan to coach, whether that behavior is a strength to reinforce or a gap to close. For each call, identify two to three direct transcript quotes. Note the timestamp, the criterion they illustrate, and the score that moment produced. Limit your session to three to five total moments across all selected calls. More than five moments is too much for an agent to process and act on. Avoid this common mistake: pulling calls to find everything wrong with an agent's performance. Effective transcript-based sessions target one to two behaviors. A manager who arrives with twelve flagged moments is running a performance review, not a coaching conversation. Insight7 links every QA criterion score to the exact quote and timestamp in the transcript. Managers can filter by criterion, identify calls where a specific behavior scored lowest, and build session preparation from pre-surfaced evidence rather than listening through hours of recordings. Step 2: Open With the Transcript Evidence, Not the Conclusion Most managers open with the conclusion: "Your empathy scores have been low." This puts the agent on the defensive before the conversation begins. Open with the evidence instead. Read the specific transcript quote, name the timestamp, and ask: "Here is what I saw at 4:32 in this call. What do you think was happening there?" This establishes that the feedback is grounded in something real and invites the agent to interpret the moment before the manager does. What Is the 70/30 Rule in Sales Coaching and Why Do New Managers Violate It? The 70/30 rule means the agent talks 70% of the time and the manager talks 30%. The manager asks questions anchored in transcript evidence rather than delivering a monologue. New managers violate this rule for a predictable reason: without prepared transcript evidence, they fill the silence with their own interpretation. Specific quotes give you material for questions: "What would you say here instead?" and "How do you think the customer interpreted this?" Those questions require the manager to say fewer words, not more. Step 3: Use Transcript Moments as Question Material During the session, every question should connect to a specific transcript moment. Instead of "how could you improve your objection handling," the question becomes: "At 7:15, the customer said they needed to think about it. You moved directly to the next talking point. What could you have said instead?" Each prepared moment generates one agent-led reflection. The manager listens and follows up. If the agent identifies the issue accurately, confirm and move on. If the agent misreads the moment, redirect with the evidence visible to both. Step 4: Annotate the Transcript Together After the agent reflects on a moment, mark up the transcript together. Write the alternative phrasing the agent identified and note which criterion that alternative would satisfy. This joint annotation converts the session from an audit into a rehearsal. The agent constructs the improvement themselves, with the original transcript as the before case. The annotated transcript becomes the accountability artifact for the follow-up session. In two weeks, when you review new calls, compare them against the annotated version. The follow-up question becomes: "Did we see this moment play out differently?" How Do You Use AI Call Summaries Effectively Without Replacing Human Coaching Judgment? AI summaries are most useful for preparation, not for the session itself. A criterion-level summary tells you which calls contain the highest and lowest scoring moments per criterion, so you can build your session plan without listening to every call in full. Which moments to address, how to sequence them, and how to respond to the agent in real time remains entirely with the manager. AI surfaces the evidence. The coaching is still human. Insight7 generates criterion-level summaries across multiple calls per agent, showing which criteria are consistently below threshold. This reduces the 60 to 90 minutes a manager would spend listening to calls before a session to a 15-minute review of pre-surfaced evidence. Step 5: Set One Behavioral Target With a Specific Criterion At the end of the session, commit to one behavioral target. Name the criterion, name the behavior, and agree on what "improved" looks like in transcript terms: "In your next two weeks of calls, when a customer raises a price objection, the
Using Self-Assessment and Recorded Interviews to Guide Coaching
Self-assessment and recorded interviews surface different types of coaching signal. Self-assessment reveals how a rep perceives their own performance. Recorded interviews reveal how that performance actually looks from the outside. The gap between the two is where the most productive coaching conversations start. This guide covers how to combine both methods systematically to guide coaching decisions. Why the Combination Matters Self-assessment alone produces coaching plans built on the rep's perception of their weaknesses, which is often inaccurate. Reps who are struggling with objection handling frequently identify their problem as "closing" because that is the point where conversations fall apart. The recorded interview shows that the real issue started three minutes earlier when they failed to acknowledge the objection before pivoting. Recorded interview review alone produces coaching plans that managers own, not reps. When managers identify problems without the rep's self-assessment as context, the rep receives feedback rather than participating in a diagnostic. Feedback compliance is lower than feedback generated through shared discovery. What are the AI personality assessment tools that integrate with coaching programs? The most commonly integrated tools are behavioral assessments (DISC, Enneagram, CliftonStrengths) and skills-based assessments (communication style, objection handling, active listening). Platforms like Cloverleaf surface DISC and Enneagram data as coaching nudges in daily workflows. For call-based coaching, Insight7 generates skills-based assessments from actual recorded calls rather than survey responses, which produces behavior evidence rather than self-report data. According to Personality Assessments for Coaching research from CoachVox, the most effective coaching integrations combine assessment data with observable behavior evidence to create coaching plans that reps recognize as accurate. Step 1: Run the Self-Assessment Before Reviewing the Recording The sequence matters. If the rep sees the recording first, their self-assessment will be anchored to what they observed rather than their genuine perception. Run the self-assessment immediately after a call session, before any review. The self-assessment should cover three questions. First, what went well in this conversation? Second, where did you feel the conversation lose momentum? Third, what would you change if you ran this conversation again? These questions surface the rep's mental model of the call before any external data shapes it. The answers create a comparison baseline for the recording review. Step 2: Review the Recording with Criteria-Mapped Timestamps Recording review without structure produces impressionistic feedback. The rep and manager watch the conversation, notice things that stand out, and discuss them. This misses patterns that are not perceptually salient but are analytically significant. Use a structured rubric that maps criteria to the call segments where they are most observable. If your rubric includes objection acknowledgment, review the segments immediately following an expressed objection. If your rubric includes discovery question depth, review the first third of the call. Insight7 connects criterion-level scores to the exact quote and call location where each score was assigned. This eliminates the review burden of watching the full recording and focuses the coaching conversation on the specific moments where criteria passed or failed. What is the most used personality assessment in sales and coaching contexts? DISC is the most commonly deployed behavioral assessment in sales and contact center coaching contexts, followed by CliftonStrengths for leadership and team development. DISC maps to call behaviors in ways that make it useful for coaching: high-D profiles tend to pivot to closing too quickly, high-S profiles struggle with urgency creation, high-C profiles over-explain before confirming interest. These patterns are observable in recorded calls and can be calibrated against self-assessment responses. The limitation of personality assessments in call coaching is that they explain tendencies, not skills. A high-D profile who has learned to slow down on objections will not behave like a high-D profile on recorded calls. Skills-based call assessment from actual recordings is more predictive of current behavior than personality type. Step 3: Compare Self-Assessment Against Call Evidence After running the self-assessment and the recorded review, place both sources of data side by side. Look for three types of gaps. Overestimation: The rep assessed their performance as strong on a criterion that the recording shows failed. This is the most common gap and requires direct evidence-based coaching. Show the specific call moment, explain why the criterion failed, and run a practice scenario targeting that behavior. Underestimation: The rep assessed their performance as weak on a criterion that the recording shows passed. This is less common but important: reps who underestimate their own competence under-deploy effective behaviors because they do not recognize them as skills. Reinforce these moments explicitly. Accurate assessment: The rep identified the same problem the recording confirms. This alignment is the foundation for intrinsic motivation to change. When the rep already knows what needs to change, coaching accelerates. Step 4: Generate Practice from the Gap Analysis The coaching plan follows from the gap analysis, not from a generic training library. TripleTen processes 6,000+ learning coach calls per month through Insight7, using the platform to identify specific performance gaps and generate targeted practice scenarios rather than assigning generic training modules. Insight7's AI coaching module generates practice scenarios from real call segments, including the specific objection types or customer personas that surfaced in the gap analysis. Fresh Prints expanded from QA to AI coaching and found that reps could practice on a specific weakness identified in their scorecard immediately after the coaching conversation rather than waiting for the next training cycle. Score tracking over unlimited retakes shows whether the practice is closing the gap. A rep who improves from 40 to 80 on an objection-handling criterion across five practice sessions has demonstrated behavior change. A rep who stays flat across five sessions needs a different coaching approach, not more of the same practice. If/Then Decision Framework If a rep overestimates performance on a specific criterion, then use recorded call evidence first before assigning practice, because the rep needs to recognize the gap before they will invest in closing it. If a rep underestimates a skill they actually demonstrate well, then reinforce that specific behavior with call evidence before assigning any additional practice, because recognition of competence
Coaching Sales Engineers Based on Technical Presentation Reviews
Sales engineers are among the least-coached people in a revenue organization, despite the fact that their demos and technical presentations are directly tied to win rates. This guide is for sales engineering managers who want to use recorded technical presentation reviews to close the coaching gap, using the same structured approach that sales managers apply to discovery calls. The underlying problem is that most SE coaching is informal: a manager sits in on a demo, gives verbal feedback, and moves on. There is no scorecard, no pattern recognition across the team, and no way to measure whether coaching is working. Recorded technical presentations change that, but only if the review process is structured around SE-specific behaviors, not generic sales criteria. What You Need Before You Start You need at least 30 recorded demo sessions from the past 60 days across your full SE team. Zoom, Google Meet, or Microsoft Teams recordings all work. You also need a list of your last 20 to 30 won and lost deals that included a technical presentation. Win-loss data connects behaviors to outcomes rather than to a manager's intuition about what good looks like. What makes a good technical sales engineer in a live demo? The behaviors that distinguish top-performing SEs are not the same as top-performing AEs. According to CloudShare's research on technical sales presentations, the behaviors most correlated with demo success include clarity of explanation for non-technical stakeholders, ability to redirect the conversation when technical objections arise, and calibration of depth to match the audience's technical sophistication. These are coachable behaviors, but they are rarely evaluated systematically. Step 1 — Define the Five to Six Behaviors That Distinguish Top-Performing SEs Start with your won deals from the past six months. Review recordings from technical presentations in those deals and ask: what did the SE do that you would want to see again? Common top-performer behaviors include: explaining architecture simply for a non-technical buyer, handling "how does this compare to our current tool?" without losing the technical stakeholder, matching feature depth to the audience's role, and adjusting language when a prospect shows confusion. Write these as observable behaviors, not attributes. "Explains clearly" is not scorable. "Uses a business analogy before going into technical architecture within the first five minutes" is scorable. Common mistake: Using your existing sales call scorecard for SE demos. A rep who scores 90% on a sales scorecard may score 60% on an SE-specific rubric because the behaviors differ. Sales scorecards evaluate rapport and closing signals, not technical clarity or audience calibration. How do you coach sales engineers based on recorded technical presentations? Review recordings against a five to six criterion rubric designed for SE interactions, not sales calls. Score each criterion on a 1 to 5 scale with behavioral anchors at each level. Identify the two to three criteria where the gap between top SEs and average SEs is widest. Build coaching sessions around those gaps using specific clips from recordings as evidence. Insight7 allows configurable criteria per call type, so an SE demo scorecard can be built separately from the sales discovery scorecard and applied to the right calls automatically. Step 2 — Score Demo Recordings Against SE-Specific Criteria Apply your criteria to a sample of 30 recordings across your full SE team. Use a 1 to 5 scale: 1 means the behavior was absent, 3 means present but inconsistent, 5 means deliberate and effective throughout. Anchor each level with a behavioral description. For "technical clarity for non-technical buyers," a 1 is "presented architecture-level detail to a business stakeholder without translation," and a 5 is "used a business analogy before each technical concept and checked for comprehension." Score recordings independently before discussing with the SE. Target inter-rater reliability above 80% if two reviewers are evaluating the same calls. Decision point: Manual scoring of 30 recordings takes 15 to 20 hours and is not repeatable at scale. Insight7 scores SE interactions using configurable criteria tuned to SE-specific behaviors, allowing you to score every demo session consistently without manual review. Step 3 — Identify the Gap Between Top SE and Average SE on Each Criterion Calculate average scores for your top-quartile SEs (top 25% by win rate) and your average SEs on each criterion. A team where top SEs score 4.2 on "stakeholder read" and average SEs score 2.1 has a large, addressable gap. A gap of 0.3 points indicates that criterion is not a meaningful differentiator. Look also for criteria where all SEs score similarly but scores are low across the board. That signals a team-wide gap: a process or product training issue, not an individual coaching issue. Individual coaching addresses individual gaps. Team-wide gaps require a different intervention. How Insight7 handles this step: Insight7's criterion-level scoring shows dimension breakdowns per SE, per team, and over time. Managers can see whether "stakeholder read" scores are improving across the team after coaching without manually reviewing calls. The evidence layer, where every score links to the exact transcript quote, means coaching conversations start from a shared factual basis rather than from a manager's memory of a session they may have observed weeks earlier. See how this works at insight7.io/improve-coaching-training. Step 4 — Build Coaching Around the Top Two to Three Gaps For each of the two to three largest gaps, find two recordings: one where a top SE demonstrates the behavior clearly, and one where an average SE fails to demonstrate it. Use these as coaching evidence. Structure each session in 30 minutes: five minutes reviewing the scorecard, 15 minutes reviewing two clips, and 10 minutes on a specific practice commitment. "Be clearer" is not a practice commitment. "Use a business analogy before explaining the integration architecture in the next demo" is. According to research from the Sales Management Association on sales coaching effectiveness, sessions that include specific behavioral evidence from recorded interactions produce measurably stronger improvement than sessions based on general feedback. Common mistake: Coaching all six criteria in one session. SEs absorb and act on two to three
Coaching Workshop Moderators on Speech Flow
Speech flow in coaching and workshop facilitation is not about speaking smoothly. It is about calibrating the pacing, emotional register, and conversational structure of a session in real time so that participants stay cognitively engaged. Analytics-informed coaching on speech flow makes this calibration observable, measurable, and improvable. What Speech Flow Analytics Actually Measures Traditional facilitator training relies on trainer observation and self-assessment. Both have systematic blind spots. Observer bias shapes what gets flagged. Self-assessment is unreliable because speakers cannot monitor their own delivery while simultaneously managing content. Speech flow analytics measures the dimensions of delivery that predict engagement: pace variation, pause frequency and duration, filler word density, sentiment arc across the session, and tonal range. These signals, drawn from recorded coaching sessions or workshops, surface patterns that neither the facilitator nor an observer would reliably detect. Insight7 analyzes call and session recordings against configurable criteria, with each criterion scored against a definition of what good and poor look like. Applied to facilitator speech, this produces a per-session scorecard showing which delivery dimensions are above benchmark and which need development. What is the flow model of coaching? The flow model of coaching applies Csikszentmihalyi's flow state theory to coaching conversations. It positions the optimal coaching interaction in the zone between challenge and skill, where the facilitator's questions and pacing are difficult enough to generate active engagement but not so demanding that the participant withdraws. Analytics-informed coaching on speech flow operationalizes this model by measuring whether the session's pacing, tonal variation, and pause structure are creating conditions for participant engagement or conditions for passive listening. Step 1: Establish a Speech Baseline From Session Recordings Record three to five representative coaching sessions or workshop segments. Score them against speech flow criteria: words per minute, pause ratio, filler word frequency, sentiment arc (does emotional tone build toward a conclusion or remain flat?), and question-to-statement ratio. These five dimensions produce a facilitator baseline. Decision point: The criterion that shows the widest variance across sessions is the highest-priority coaching target. A facilitator who delivers consistent pacing but wildly variable question frequency is getting inconsistent participant engagement in ways that correlate with the question variation, not the pacing. Common mistake: Starting coaching feedback with overall delivery ratings. Generic ratings ("you need more pauses") do not change behavior. Moment-specific feedback does: "at minute 14, your words-per-minute jumped from 140 to 185 and the participant response rate dropped." According to the Association for Talent Development's research on coaching effectiveness, specific and timely feedback tied to observable behaviors produces stronger skill improvement than delayed, general performance ratings. Step 2: Map Analytics to Specific Session Moments Generic feedback does not create the behavioral specificity needed for improvement. The coaching recommendation must name the moment, the behavior, and the participant outcome. How to build moment-specific coaching from analytics: Pull the timestamp where pace variation increased sharply. Review what was happening at that point: topic transition, difficult question, participant pushback? Note whether participant response rate changed within two minutes of the pace shift. Build the coaching recommendation around that specific moment, not a general delivery summary. Insight7's post-session AI coaching provides voice-based interactive reflection. Rather than just delivering a score, it engages the facilitator in a discussion about what happened at specific moments in the session, creating the mechanism for deliberate practice rather than generic awareness. Specific thresholds to track: Words per minute above 175 in a coaching session reduces participant uptake, as comprehension research consistently shows. Pause ratio below 8% of session time correlates with reduced participant processing of complex questions. Filler word density above 3 per minute signals preparation gaps rather than delivery style. Step 3: Apply Mood Analytics to Session Design Decisions Mood analytics in coaching sessions measures the emotional register of the facilitator and the sentiment arc across the session. The signal is whether tone stays flat, builds progressively, or spikes and drops. Each pattern produces a different participant outcome. Flat sentiment arc: Consistent moderate tone throughout. Participants remain engaged but rarely reach insight moments. Coaching recommendation: introduce deliberate tonal variation at the 30 and 60 percent marks. Spike-and-drop pattern: High emotional engagement in the opening, falling to neutral mid-session. Common in facilitators who front-load energy. Coaching recommendation: pace the high-energy moments across the session rather than concentrating them at the start. Progressive build: Emotional register rises progressively toward the conclusion. Highest correlation with participant-reported insight. Coaching recommendation: study the delivery behaviors in sessions where this occurs and replicate them. Insight7 extracts tone analysis from session recordings, evaluating sentiment and tonality beyond transcripts. Applied to facilitator coaching, this surfaces which session structures produce which mood arcs and which delivery behaviors drive participant engagement. What are the 22 flow triggers? The 22 flow triggers, documented by researcher Steven Kotler and the Flow Research Collective, fall into four categories: psychological (clear goals, immediate feedback, challenge-skill balance, undivided focus), environmental (high consequences, rich environment, deep embodiment), social (serious concentration, shared risk, close listening, autonomy, familiarity), and creative (creativity, risk, complexity, unpredictability, novelty). For workshop facilitators, the most directly applicable triggers are immediate feedback and challenge-skill balance, both of which can be monitored and coached using speech flow analytics. If/Then Decision Framework If a facilitator's speech analytics show high words-per-minute with low pause ratio, then coach specifically on pause placement after questions, because unpacking time is what converts questions into engagement rather than passive reception. If mood analytics show a spike-and-drop sentiment arc, then map the session timeline to identify where the facilitator's energy is concentrated and redistribute it, because front-loading energy depletes engagement for the insight-generating moments mid-session. If criterion-level feedback is not producing behavior change after three sessions, then shift from score delivery to moment-specific feedback tied to session timestamps, because generic ratings do not create the behavioral specificity needed for change. If facilitators are resistant to analytics-informed feedback, then start with self-comparison data (this session versus the facilitator's own baseline) rather than benchmark comparisons, because self-referenced improvement is less confrontational and produces stronger adoption. If you need to connect
How to Automate Coaching Evaluation Templates with AI
QA managers and coaching program leads spend hours each week building session plans, pulling performance data, and formatting evaluation documents, time that could go toward the actual coaching conversation. AI tools that automate coaching session plans, documents, and templates cut that administrative load and produce more consistent coaching records in the process. Why Do Manual Coaching Templates Create Inconsistency Across Programs? ATD research on workplace coaching programs finds that documentation quality varies widely when managers build evaluation templates from scratch, leading to coaching records that reflect individual manager habits more than actual rep performance. When each team lead formats a session plan differently, program leads have no clean way to aggregate coaching data, spot trends, or demonstrate ROI to leadership. Standardizing templates is the first step, and automating their population from real performance data is what makes standardization scalable. Step 1: Define Your Coaching Framework Before Touching Any Tool Automation only works if there is a framework to automate. Before configuring any platform, document: The dimensions your coaching program evaluates (tone, resolution rate, compliance, objection handling) The scoring scale for each dimension The structure of a standard session plan: prep review, call examples, agreed focus areas, follow-up actions The cadence: weekly one-on-ones, bi-weekly group coaching, monthly calibration reviews This framework becomes the schema that AI tools populate. If the framework is undefined, the tool produces filled-out templates that are structurally inconsistent, which defeats the purpose. Step 2: Connect Performance Data to Your Evaluation Input Coaching evaluation templates need data inputs to be useful. The two main sources are QA scores from call analysis and self-reported rep goals from your performance management system. Insight7 analyzes 100% of call recordings automatically, generating per-rep score breakdowns by the dimensions on your coaching scorecard. Instead of a QA analyst manually reviewing calls before each session, the platform surfaces the relevant call data, flags the sessions most worth reviewing, and formats the output around the focus areas your framework defines. Connect Insight7 to your coaching workflow using these setup steps: Map your existing QA scorecard dimensions to Insight7's configurable rubric fields Set the review window for each coaching cycle (weekly, bi-weekly) Enable automated rep summaries so each session plan pre-populates with the rep's score trends, top-performing behaviors, and priority development areas Export or integrate directly into the document layer where your coaching records live Step 3: Build Template Automation in Your Performance Platform QA score data feeds the evaluation, but the session plan document itself needs a home. Performance management platforms handle the scheduling, template structure, and record-keeping side of coaching automation. Lattice supports customizable 1:1 templates with structured talking points, action item tracking, and manager prep prompts. You can build your coaching framework directly into the template so every session follows the same structure, regardless of which manager runs it. 15Five adds a check-in workflow that prompts reps to self-assess on the same dimensions your QA scorecard uses before each session, giving managers a pre-populated starting point without any manual prep. Leapsome integrates learning goals with coaching records, which is useful for coaching programs that tie session outcomes to skill development tracks. The right choice depends on whether your coaching program is primarily manager-driven (Lattice works well), rep-driven with self-assessment (15Five fits better), or integrated with a broader learning and development function (Leapsome is worth evaluating). Step 4: Automate Pre-Session Document Generation The most time-consuming part of coaching preparation is pulling together the relevant calls, scores, and context before the session starts. Automate this with a triggered workflow that runs 24 to 48 hours before each scheduled session. Using Insight7's QA and reporting layer, set up a pre-session export that includes: The rep's QA score trend over the coaching window The two or three calls most relevant to the session's focus areas (one strong example, one development opportunity) A summary of the dimensions where the rep improved versus the prior period Any compliance flags from the period Push this export into the session plan template in your performance platform. The manager opens the meeting with a fully prepared document rather than spending 20 minutes the morning of the session pulling data from multiple systems. Step 5: Standardize Post-Session Documentation Coaching programs lose continuity when post-session notes are unstructured. Managers write different things in different places, agreed actions are not tracked, and the next session starts without a clear read on what happened last time. Build a post-session template that captures four fields: the session's focus area, what the call examples showed, the rep's agreed development action for the next period, and the manager's follow-up commitment. Lock these fields in your performance platform so they are required before the session record closes. Lattice and Leapsome both support required-field enforcement on 1:1 templates. After each session cycle, Insight7's rep-level trend data shows whether the behaviors addressed in coaching are improving on actual calls. This closes the loop between what was discussed in the session and what is happening in the field. Step 6: Build a Coaching Calendar Tied to QA Cycle Output Coaching programs are most effective when session timing aligns with the QA review cycle. If QA data updates weekly, weekly coaching sessions can use current data. If QA runs bi-weekly, coaching cadence should match. Map your Insight7 analysis schedule to your coaching calendar in your performance platform. Set automated reminders that trigger when a new QA summary is ready for a given rep, prompting the manager to schedule or prep for the next session. This turns coaching from a calendar obligation into a data-triggered workflow. How Do You Measure the ROI of Automated Coaching Templates? SHRM research on performance management programs identifies documentation consistency and follow-through on agreed actions as the two strongest predictors of coaching program effectiveness. Operationally, track: percentage of scheduled sessions that have a completed pre-session document (target above 90%), percentage of post-session records with all required fields completed, and the correlation between coaching session frequency and QA score improvement per rep over 90 days. If the automation is working,
Scoring Real-Time Coaching Calls for Verbal Effectiveness
Contact center supervisors who rely on manager intuition to evaluate coaching call quality end up with inconsistent feedback, contested scores, and agents who do not know what to work on. Scoring coaching calls systematically for verbal effectiveness creates an objective baseline that connects practice performance to live call improvement. This guide covers how to score coaching calls for verbal effectiveness, which criteria matter most for customer service agents, and how to use those scores to build a sustainable coaching cycle. Why Scoring Verbal Effectiveness Matters for Customer Service Agents Verbal effectiveness in customer service covers the behaviors that determine whether a customer feels heard and helped: empathy acknowledgment, clear explanation of next steps, confidence under pressure, and de-escalation language. These are scoreable. They are not inherently subjective, but they are treated as subjective when teams lack a defined rubric. Insight7 evaluates verbal behaviors through both intent-based and script-based criteria, depending on what the team defines as "good." An empathy check can be evaluated on whether the agent said the exact phrase, or on whether the intent of the phrase was achieved in the conversation. This distinction matters because verbatim compliance scoring penalizes agents who use natural language that achieves the same result. How to effectively coach call center agents? Effective call center coaching follows a four-step loop: score calls using defined criteria, identify the one to two behaviors each agent needs to improve, assign targeted practice scenarios built from real failing calls, and re-score live calls after training to confirm the behavior changed. The most common error is coaching from aggregate scores rather than criterion-level breakdowns. An agent at 69% overall could be performing at 85% on compliance and 45% on empathy. A general coaching session does nothing for them. A targeted empathy scenario built from their specific failing calls does. Insight7 delivers criterion-level scoring tied to transcript evidence for every call. What is the 70-30 rule in coaching? The 70-30 coaching principle suggests that the person being coached should do 70% of the talking while the coach does 30%. In call center coaching, this translates to asking the agent to diagnose their own call before the coach provides feedback. "What do you think happened at the 3-minute mark?" produces more durable learning than "at the 3-minute mark, you should have acknowledged the customer's concern before moving to the resolution." Evidence-based scoring supports this approach: when a score is linked to a specific transcript quote, the agent can see exactly what the score is evaluating, making self-diagnosis accurate rather than defensive. Steps for Scoring Coaching Calls for Verbal Effectiveness Step 1: Define your verbal effectiveness criteria with behavioral anchors. Generic criteria produce generic scores. Define each criterion with a specific behavioral anchor: what does "good" empathy sound like in your context, what does "poor" empathy sound like. Empathy without anchors will be scored inconsistently across raters and across time. Anchors for empathy might include: good = agent names the customer's specific concern before offering a solution; poor = agent moves to resolution without acknowledging the stated concern. Insight7 stores these anchors in a "context" column per criterion. Calibration typically takes 4-6 weeks to align automated scoring with human judgment. Step 2: Score 100% of coaching sessions, not a sample. Common mistake: treating coaching call scoring as optional or sampling-based. If you score 1 in 5 coaching sessions, you cannot detect whether a rep is improving or whether a particular scenario type is too easy. Full coverage gives you a reliable trend line. Insight7 processes roleplay sessions automatically after completion, generating a per-session scorecard that shows performance on each criterion and flags improvement or regression. Step 3: Compare coaching session scores to live call scores on the same criteria. This is the validation step most teams skip. A rep can score 82% on an empathy criterion in a roleplay scenario and still score 58% on the same criterion in live calls. The gap between coaching performance and live performance tells you two things: (1) whether the scenario is realistic enough to transfer, and (2) whether the rep is applying the learned behavior under real pressure. Pull the same criteria from live call scoring and coaching session scoring for a side-by-side comparison. Insight7's per-agent scorecard shows both views in one dashboard. Step 4: Set a passing threshold before each session, not after. A threshold defined after the fact is not a standard, it is a rationalization. Before assigning a coaching scenario, define: this rep must score 75 or above on empathy criteria in three consecutive sessions before we consider the behavior embedded. Reps who know the threshold can track their own progress. ATD research on deliberate practice shows that learners who know their target score before a practice session improve significantly faster than those who receive scores without context. Step 5: Use failing verbal effectiveness calls as scenario source material. The most effective roleplay scenarios come from the calls where verbal effectiveness failed most clearly: the escalation call where the agent skipped empathy and went straight to policy recitation, the complaint call where tone became clipped under pressure. Insight7's AI coaching module converts real call transcripts into scenarios with configurable personas that replicate the emotional tone and communication style of the original interaction. Agents practice the specific situation they struggled with, not a generic difficult customer scenario. Decision point: if no specific failing calls exist, use top-performer calls as positive models rather than invented scenarios. If/Then Decision Framework If agent scores vary widely call to call → then the criteria need behavioral anchors before scoring is reliable. If coaching session scores are high but live call scores are not improving → then the scenarios need to be made harder to match real call pressure. If multiple agents fail on the same verbal effectiveness criterion → then the issue is systemic and needs a team training session before individual coaching. If a rep plateaus after five coaching sessions with no improvement → then escalate to a live coaching conversation rather than continuing
Reviewing 1:1 Sales Coaching Calls to Drive Rep Improvement
Sales rep performance tools have split into two categories: platforms that score calls and flag gaps, and platforms that motivate reps to close those gaps through competitive mechanics. The best sales coaching tools with leaderboards, missions, and coaching tips combine both layers, giving managers data to coach from and giving reps a visible reason to improve. This guide covers sales rep tools with gamification features, evaluated for sales managers who need both the performance analytics and the rep engagement mechanics to make coaching programs actually stick. If/Then Decision Framework What is the best sales rep tool with leaderboards and coaching missions? The best tool depends on whether your coaching program is built around call performance data or CRM activity metrics. For sales teams where conversation quality drives outcomes, Insight7 connects leaderboard positions to actual call criterion scores. For teams where activity volume is the primary driver, Ambition or SalesCompete deliver stronger gamification mechanics against CRM data. If your sales team needs leaderboards tied to call behavior metrics (objection handling, discovery depth, compliance), then use Insight7, because criterion-level scoring from 100% of calls gives leaderboards statistical validity that activity-based boards lack. If you need enterprise gamification with missions, competitions, and TV leaderboards for a large floor, then use Ambition, because their scorecards and competition mechanics are the most configurable in the market for high-volume sales environments. If your coaching program needs AI-generated missions based on skill gaps rather than manager-assigned challenges, then use Mindtickle, because the readiness scoring layer generates missions targeted at each rep's specific development gap. If you manage a field sales team and need territory-level leaderboards alongside coaching, then use SPOTIO, because field activity tracking integrates natively with their leaderboard mechanics in a way that office-focused platforms do not replicate. If your team runs Slack and you need lightweight gamification without a separate platform, then use SalesCompete, because their CRM-connected Slack competitions run without asking reps to log into another tool. If your coaching focus is practice session completion and skill improvement, then use Second Nature, because scenario completion scores and session-level leaderboards tie gamification directly to practiced behaviors. Ambition Ambition is the most recognized enterprise gamification platform for sales teams. Managers configure scorecards combining revenue metrics, activity data, and CRM event triggers. Competitions run at the team, pod, or individual level. TV display dashboards make leaderboard positions visible on the sales floor. The platform does not generate coaching tips from call data. Coaching is manager-driven: Ambition surfaces performance data, managers translate it into coaching actions. Best suited for high-volume inside sales floors where visibility and competition drive activity. Pro: The most configurable gamification mechanics in the market for enterprise sales. Competitions can target any CRM metric without custom development. Con: No AI coaching tips generated from call data. Coaching quality depends entirely on manager skill and engagement. Pricing: Enterprise pricing, contact vendor. When the goal is motivating high-volume activity through competition rather than improving conversation quality, Ambition is the category leader. Insight7 Insight7 connects sales coaching to call performance data at the criterion level. The QA engine scores 100% of recorded calls on configurable criteria: discovery depth, objection handling, compliance language, empathy, and process adherence. Leaderboards rank reps on these behavioral dimensions, not just revenue or activity. Practice scenarios are generated from the hardest real calls in the data. The platform auto-suggests roleplay missions based on QA findings. Managers approve before assignment, keeping human judgment in the loop. Reps retake scenarios unlimited times with score tracking showing improvement trajectory. Pro: The only platform in this list that connects leaderboard position to specific call behaviors reps can practice. Missions are generated from live call data, not manager intuition. Con: Requires an existing call recording infrastructure. Gamification mechanics are not as configurable as Ambition for pure competition scenarios. Pricing: AI coaching from $9/user/month at scale. See insight7.io/pricing. For sales managers who want leaderboards and missions to drive actual conversation skill improvement rather than activity volume, Insight7 is the only platform that closes the loop from call data to practice assignment. Mindtickle Mindtickle integrates training content, AI roleplay simulation, and sales gamification in one platform. The readiness score combines knowledge assessment, practice session performance, and activity data into a single rep score. Missions target the lowest-scoring readiness dimensions per rep. Leaderboards show readiness scores alongside activity metrics. Managers can run challenges on specific skills or content completion. Best suited for enterprise sales organizations with structured onboarding where training completion drives readiness tracking. Pro: AI-generated missions targeted to each rep's specific skill gaps are the most sophisticated mission-assignment mechanism in this list for training-based programs. Con: Readiness scoring is based on training content, not live call performance. Reps who complete training but underperform on live calls show green without the real-world evidence. Pricing: Enterprise pricing, contact vendor. The readiness score combining multiple data types gives sales managers a single metric that Ambition's activity-only approach cannot replicate. SPOTIO SPOTIO is a field sales performance platform with territory management, leaderboards, and coaching features. Leaderboards are built around field activity: visits made, deals advanced, territory coverage. Managers see which reps are working which accounts and how activity translates to pipeline. Coaching in SPOTIO is activity-based. Managers review field activity data and provide coaching direction. There is no AI-generated coaching from call transcript data. Pro: The only platform in this list built specifically for field sales activity tracking. Territory coverage data gives managers coaching context that office-focused platforms cannot provide. Con: Limited to field activity data. No conversation quality scoring or call-based coaching tips. Pricing: Contact vendor for current pricing. When reps spend most of their time in the field rather than on calls, SPOTIO's territory leaderboard format is the only option that reflects how their work is actually structured. SalesCompete SalesCompete runs sales competitions and leaderboards through Slack using CRM event data. When a rep closes a deal, logs a meeting, or hits a milestone, SalesCompete posts the achievement and updates leaderboard rankings in Slack channels. Competitions are configurable: first to a target,
The Must-Have Fields in Your CX Coaching Template
CX managers who run 1:1 coaching sessions without a structured template spend the first ten minutes of every meeting figuring out what to talk about. A good coaching template fixes that by capturing the right data before the conversation starts and giving both manager and agent a shared structure for what comes next. This guide walks through the seven fields every CX coaching template needs, and shows where AI tools can pre-populate the data-heavy sections so managers spend their time on actual coaching rather than call retrieval. Methodology A 1:1 coaching template should do two things: anchor the conversation in evidence from real calls, and convert that evidence into a specific commitment from the agent. The seven fields below move in sequence from evidence to diagnosis to action. Fields 1 through 3 are data-entry tasks that can be automated. Fields 4 through 7 require manager judgment and cannot be automated. What is a good 1 on 1 agenda? A good 1:1 coaching agenda follows a simple logic: start with what happened (evidence from calls), move to why it happened (root cause), then finish with what changes (commitment and follow-up). The most common mistake is spending the session on the evidence phase because the manager had to pull and review calls manually before or during the meeting. When fields 1 through 3 are pre-populated from your QA system, the conversation can start at the diagnosis stage, which is where coaching actually happens. Field 1: Call and Interaction Reference Every coaching session should be tied to specific calls, not a general impression of the agent's recent performance. This field captures the date of the call, the call ID or recording link, and any relevant context (inbound vs. outbound, call type, queue). What to include: Date and time of the call Call ID or direct link to the recording Call type and queue context QA evaluation date if different from call date Without a specific call reference, agents cannot replay the moment being discussed and managers cannot verify their own recall. General impressions of "how you've been doing lately" are not coachable. Specific calls are. Insight7 populates this field automatically. Every scored call is stored with its transcript, recording link, and metadata. Managers open the agent scorecard, select the calls they want to discuss, and the reference data is already there. Field 2: Criteria Scores with Evidence This field captures the numerical or weighted score on each evaluation criterion, and the specific transcript quote or interaction moment that supports it. The score alone is not enough. An agent who scored 60% on empathy cannot improve without knowing exactly which moment in which call produced that result. What to include: Score per criterion (on your standard scorecard) Direct transcript quote or timestamped recording clip Whether the score was above or below the agent's average Avoid this common mistake: recording only summary scores without evidence. "Empathy: 3/5" tells an agent what their score is but not what behavior created it. The coaching conversation stalls because the agent has no reference point for what they did differently in this call versus a higher-scoring one. Insight7 links every criterion score to the exact quote and location in the transcript. Managers can click through to verify any score, and agents can see the same evidence during the 1:1. This replaces a 15-minute call review exercise with a 30-second reference lookup. Field 3: Behavioral Gap Description A behavioral gap is the difference between what the agent did and what good performance looks like for that criterion. This is distinct from the score: the score tells you the gap exists; the behavioral description tells you what the gap looks like in practice. What to include: What the agent did (tied to the transcript evidence from Field 2) What good performance looks like on this criterion Whether the gap is a pattern across multiple calls or an isolated instance Good behavioral gap descriptions are specific and observable. "You interrupted the customer three times in the first two minutes" is coachable. "Your listening skills need work" is not. Insight7 pre-generates criterion context based on how your scorecard is configured, including descriptions of what good and poor look like for each item. When a score falls below threshold, the system surfaces the relevant gap description alongside the evidence. Managers can edit or add context before the 1:1. What is a customer experience playbook? A customer experience playbook is a set of documented behaviors, scripts, and decision rules that define how agents should handle each interaction type. A CX coaching template is the session-level tool that connects actual call data to the behaviors defined in the playbook. Without call evidence (Fields 1 through 3), a coaching session is just a re-read of the playbook. With evidence, it is a targeted intervention tied to a specific deviation from the expected behavior. Field 4: Root Cause This is the first field that requires manager judgment. Root cause identifies why the behavioral gap exists. The three categories most relevant to contact center coaching are: Skill gap: The agent knows what good looks like but cannot execute it consistently. Needs practice. Knowledge gap: The agent does not know the correct behavior, policy, or product information. Needs training or information. Motivation or environment factor: The agent knows and is capable but is not applying the behavior. Needs a different conversation. Getting root cause wrong leads to the wrong intervention. An agent with a knowledge gap who gets assigned a role-play practice session will practice the wrong behavior. An agent with a skill gap who gets sent to a training module will not develop the muscle memory needed to change their call behavior. Field 5: Coaching Assignment Based on the root cause identified in Field 4, this field records the specific action the agent will take before the next 1:1. The assignment should be concrete and verifiable. Examples by root cause: Skill gap: Complete two role-play practice sessions on [objection handling] using the AI coaching module and retake until
What to Include in a Sales Coaching Template for Role-Based Skills
A sales coaching template for cold calling and outreach fails when it tries to cover every role with the same framework. An SDR making 80 outbound dials a day has different coaching needs than a field rep doing 10 strategic calls a week. A template that ignores this produces generic feedback that doesn't change behavior. This guide covers what to include in a sales coaching plan for cold calling, organized by role and skill area, with specific criteria and how to use call data to keep the plan calibrated over time. What a Sales Coaching Plan for Cold Calling Needs to Cover Most templates address the wrong level of detail. They list competencies like "prospecting skills" without specifying what observable behavior in a cold call demonstrates those competencies. A useful coaching plan defines: (1) the specific behaviors to observe during the call, (2) what "good" looks like for each behavior at this role level, (3) how to score each criterion consistently across coaches, and (4) how to connect scorecard feedback to targeted practice. What should a good opening line in a cold call accomplish? A strong cold call opener does two things in the first 15 seconds: establishes specific relevance (why this rep is calling this person at this company right now) and earns permission to continue. Generic openers like "I wanted to introduce myself" fail the relevance test. Openers that name a specific problem or outcome the prospect is likely facing, without reading from a script, consistently outperform in call analysis data. According to research from RAIN Group, the ability to connect quickly to a buyer priority is among the top differentiators between top and average cold callers. Sales Hacker research on cold call structures similarly identifies the opener as the highest-leverage moment to coach. How long should a cold call opening last before asking a question? The first question in a cold call should come within 45 to 60 seconds of the opener. Calls where reps monologue past 90 seconds before asking a question show significantly higher disengagement and early termination rates in call analytics reviews. The opener creates the context; the first question tests relevance and earns the right to continue. Coaching templates should define the opener-to-first-question timing as a scorable criterion. Role-Based Coaching Criteria A useful coaching template segments criteria by role because observable good behaviors differ by position and call type. For SDRs focused on top-of-funnel outbound, coaching criteria center on the opener, qualification, and handoff. Evaluate whether the SDR states a specific reason for the call, names a relevant problem, and asks a qualifying question within the first 60 seconds. For AEs managing follow-up and outreach, coaching criteria shift to discovery depth and deal progression. The AE should cover budget, authority, timeline, and pain before transitioning to solution, connect specific outcomes to prospect priorities, and own next steps explicitly. Insight7's weighted criteria system allows each criterion to have a main item, sub-criteria, descriptions, and a context column defining what "good" and "poor" look like. This structure is directly applicable to cold call coaching templates because it forces specificity and reduces scoring inconsistency between managers. If/Then Decision Framework If you need to score specific cold call behaviors across a large SDR team automatically, then use Insight7 to score 100% of calls against your defined criteria. If you need to translate coaching scores into targeted practice scenarios, then use Insight7's AI coaching module to generate role-play sessions from real cold calls where specific criteria failed. If you need coaching criteria calibrated to top performer behavior on your specific call types, then run a call analysis comparing top quartile SDRs against bottom quartile on each criterion before finalizing weightings. If you need to measure whether coaching plan interventions produce score improvement over time, then use a call analytics platform with time-series scoring rather than a static scorecard spreadsheet. If the primary gap is delivery skills (filler words, pacing, energy) rather than conversation structure, then add a speech feedback tool like Yoodli alongside call scoring for delivery-specific coaching. Building the Coaching Plan: Key Sections A complete sales coaching plan for cold calling and outreach includes these components. Role definition and call objectives. Define what a successful call looks like in outcome terms. For an SDR, success is a qualified meeting booked. For an AE on a follow-up, success is a confirmed next step with a specific date. These outcome definitions anchor all downstream criteria. Observable criteria with scoring anchors. For each competency, define it in terms of observable call behavior. "Strong opener" is not a criterion. "States a specific, role-relevant reason for the call within 15 seconds without reading from a script" is a criterion. Define your scoring scale with anchors at each level. A 1-5 scale without anchors produces variance between managers. Weighted criteria reflecting cold call priorities. Not all criteria carry equal weight. For cold calling, opener quality and discovery question discipline are highest-leverage because they determine whether the call continues. Allocate 30 to 40% of total score weight to the opener and qualifying exchange. Connection to practice. For each criterion where a rep scores consistently below target, specify a practice format. Insight7's coaching module generates AI practice scenarios from real call transcripts. When a manager identifies an SDR not asking qualifying questions early enough, the system generates a scenario from real calls where this failure occurred, giving targeted practice rather than generic drills. TripleTen uses this approach across 6,000+ learning coach calls per month. Review cadence. Coaching templates drift as market conditions change and messaging evolves. Run a quarterly comparison between top and bottom performers on each criterion using call analytics. If a criterion stops differentiating, recalibrate or replace it. Scoring Criteria for a Cold Call Coaching Template Most effective templates for cold calling include 5 to 8 primary criteria with 2 to 3 sub-criteria each. Fewer than 5 misses important dimensions; more than 10 creates coaching session overload. For a 15-minute cold call, a well-structured template covers: Criterion What It Measures Suggested Weight
How to Track Sales Rep Improvement with Coaching Logs
Sales managers who implement coaching programs without a tracking system face a common problem six weeks later: they cannot tell whether the coaching worked. Reps seem to be performing better, or worse, or the same, but there is no data connecting a specific coaching session to a measurable change in call behavior. This guide walks through a six-step process for building a coaching log system that produces that data, using QA scoring to isolate improvement at the criterion level. Step 1: Define the Metrics That Coaching Targets Before logging a single session, identify which QA criteria you are trying to move. Not all metrics improve from coaching. Activity metrics like call volume are driven by scheduling, not skill. The metrics that respond to coaching are behavioral: objection handling score, discovery question frequency, value framing before pricing, compliance phrase usage. Build a list of 5-8 specific QA criteria that are both measurable in call recordings and influenced by rep behavior in real time. These become the official coaching targets. Every session in your log will reference at least one criterion from this list. Without this step, coaching logs become notes about conversations rather than records of targeted behavior change. Insight7's weighted criteria system supports this setup: managers configure main criteria, sub-criteria, and descriptions of what good and poor performance look like for each. These definitions become the shared language between manager and rep during coaching sessions. What Makes a Coaching Metric Trackable A trackable coaching metric has three properties: it appears in call recordings (so it can be scored before and after coaching), it is specific enough to define as present or absent (not "communication quality" but "stated the refund policy within the first 60 seconds"), and it is influenced by rep choice rather than by factors outside their control. Metrics that pass all three are worth tracking. Those that fail any one of these tests create noise in your improvement data. Step 2: Log Each Session with the Criterion Targeted A coaching log entry should take under two minutes to create. The minimum required fields are: rep name, session date, criterion targeted, baseline score on that criterion, coaching method used (roleplay, call review, script walkthrough), and a one-sentence description of what the rep agreed to do differently. The criterion field is the most important. Teams that log "discussed objection handling" instead of "targeted criterion: handles price objection before closing" cannot aggregate data later. When every log entry references a specific criterion from your defined list, you can run queries like "which criterion did we coach most this quarter, and what was the average score change after coaching?" Avoid this common mistake: logging the coaching topic as a broad theme rather than the specific QA criterion. "Communication" and "customer empathy" are themes. "Used empathy statement within 30 seconds of customer frustration" is a criterion. Only criteria produce trackable improvement data. Step 3: Record Baseline Scores Before the Session Before the coaching session occurs, pull the rep's current average score on the targeted criterion from the last 10 calls. This is the baseline. Record it in the log entry at the time of session scheduling, not after. Recording baseline after the fact introduces bias. Managers who know coaching went well tend to remember lower starting scores. Managers who feel the coaching did not land tend to remember higher starting scores. The baseline must be recorded from the scoring system before any subjective judgment about the session outcome enters the picture. Insight7 generates per-rep, per-criterion scorecards automatically. Before a coaching session, a manager can pull the rep's score on the targeted criterion across their last 10 calls and record the average in under a minute. This removes the manual effort of reviewing individual call recordings to establish a baseline. Step 4: Re-Score the Same Criterion in the 10 Calls After Coaching Two to three weeks after the coaching session, collect scores for the targeted criterion from the rep's next 10 calls. This is the post-coaching window. Use the same scoring tool and criteria configuration that generated the baseline, so the two numbers are directly comparable. Do not use the first call after coaching as the measurement. Reps often overcorrect immediately after a session and then revert. The 10-call window smooths out that overcorrection and captures whether the behavior change held. Platforms that score 100% of calls automatically, like Insight7, generate this data without requiring the manager to manually select and review calls. What a Valid Pre/Post Comparison Requires For a pre/post comparison to be valid, four conditions should hold: the same criterion was scored in both periods, the scoring criteria definition did not change between baseline and post window, the call types in both windows are comparable (inbound vs. inbound, not inbound vs. outbound), and the post window represents at least 10 calls. If any of these conditions are not met, the comparison may reflect criterion drift or call type variation rather than actual rep improvement. Step 5: Compare Pre/Post Scores to Isolate Improvement Calculate the score change: post-coaching average minus pre-coaching average on the targeted criterion. A positive number indicates improvement. A flat or negative number indicates the coaching approach did not produce the intended change. Isolating improvement at the criterion level is the key advantage of this system over general performance tracking. If a rep's overall QA score improves, you cannot tell whether the coaching caused it or whether unrelated factors shifted. When you track a specific criterion from before a specific coaching session to after, you have isolated the variable. SQM Group research on contact center performance improvement shows that criterion-level tracking doubles the likelihood that managers identify which interventions worked versus which did not. This is the difference between knowing coaching happened and knowing coaching worked. Step 6: Aggregate Across the Team to Identify Which Coaching Approaches Work After three months of logging, you have enough data to answer questions your coaching program has probably never answered before: which criteria show the largest average improvement after coaching, which
