HR and L&D managers often treat performance management issues and training gaps as the same problem. They are not. A rep who received thorough training and still underperforms has a performance management issue. A rep who never received clear instruction on the expected behavior has a training gap. Conflating the two leads to the wrong intervention: more training for people who need accountability, and performance improvement plans for people who need better instruction. This six-step guide uses feedback analysis to make that distinction before any intervention is designed.
SHRM's toolkit on managing employee performance notes that organizations using structured, criterion-level feedback data in performance reviews identify correctable skill gaps earlier and can more consistently distinguish training gaps from sustained performance deficiencies.
Step 1: Distinguish Performance Issues from Training Gaps Using Criterion-Level Data
The first sorting question is: did this person have access to clear, documented instruction on the behavior in question, and did the instruction produce a behavior change in any measurable period after delivery? If the answer is yes to the first and no to the second over at least 60 days, you have a performance issue. If the answer is no to the first, you have a training gap.
Pull criterion-level scores for each rep from your most recent 30-day period. Flag criteria where the rep scores below threshold, typically 75% for behavioral items. For each flagged criterion, check whether a training module or coaching session covering that criterion was completed in the prior 90 days.
Decision point: If you do not have criterion-level performance data, you cannot make this distinction accurately. Assessment through manager observation alone produces inconsistent classifications because different managers apply different standards to the same behaviors. Establish a baseline scoring system before attempting to classify performance issues.
Insight7's QA platform surfaces criterion-level performance per agent and per team. This produces the training-versus-performance sorting data you need without requiring manual call review across your full agent population.
Step 2: Build a Feedback Data Collection Framework
Effective feedback analysis requires at least three data streams: call QA scores, manager observation notes, and agent self-assessment or survey data. Each captures a different layer. QA scores capture behavioral consistency on observable criteria. Manager notes capture qualitative patterns and context. Self-assessment data reveals whether the agent perceives a gap that the manager has not formally surfaced.
Map your current data sources to these three streams. Identify which are being collected consistently and which have coverage gaps. A feedback framework that relies entirely on QA scores misses motivational and contextual factors. A framework relying entirely on manager observation introduces rater inconsistency.
Set a documentation standard for manager observation: every note should reference a specific call or interaction, name the criterion being observed, and note whether the behavior met or fell below the rubric standard. Narrative notes without behavioral anchors are not usable for pattern analysis.
Common mistake: Collecting survey data annually rather than at regular intervals tied to training delivery. Survey data that lags training by six months cannot tell you whether the training produced a behavior change. Collect self-assessment within two weeks of any major training module.
Step 3: Aggregate Feedback by Team and Role to Find Systemic Patterns
Individual feedback data tells you about individual performance. Aggregated feedback tells you whether your training program is working. Aggregate criterion-level scores by team and role for your most recent quarter. Look for criteria that fail across a majority of a team or role group. Those are systemic failures in the training program, not individual performance issues.
Insight7's analytics dashboard clusters multiple calls per rep into scorecards and shows team-level trend data by criterion. For L&D managers assessing training effectiveness, the key view is the criterion trend over time: did scores on a trained criterion improve after training delivery?
Run this aggregation quarterly at minimum. Monthly aggregation is useful for monitoring but produces too much variance to support program decisions. Set a threshold: if 40% or more of agents on a team score below the criterion target after training delivery, the training program needs redesign, not the agents.
Decision point: If your team spans multiple roles with different call types, do not aggregate across roles. Combining scores from customer service reps and sales reps obscures both patterns. Aggregate within role groups only.
Step 4: Identify the Gap Between Documented Feedback and Actual Behavior Change
Documentation of feedback delivery does not equal behavior change. Many organizations have records showing coaching sessions were completed but cannot show whether behavior changed in the 30 to 60 days following. This is the gap that performance management most commonly falls into.
For each criterion-level gap identified in Step 3, build a before-and-after comparison. Pull criterion scores from the four-week period before coaching was delivered and the four-week period after. A gap that does not close after two coaching cycles with documented delivery is a performance management issue, not a training gap.
Track this at the individual level within the aggregated team view. A team average may show improvement while a subset of agents shows no change. Those agents require individual performance plans, not group retraining.
Common mistake: Counting coaching session attendance as evidence of behavior change. Attendance is an input metric. Criterion score change is an output metric. Track outputs.
Step 5: Set Performance Improvement Thresholds Tied to Specific Criteria
Generic performance improvement plans that say "improve call quality" are not enforceable or measurable. Effective PIPs name the specific criterion, the current score, the target score, the timeline, and the intervention assigned. Each of those five elements must be present for the plan to be operationally useful.
Example: "Criterion: Call resolution confirmation. Current 30-day average: 58%. Target: 75%. Timeline: 90 days. Intervention: Weekly coached call review with manager, plus two AI coaching roleplay sessions on resolution confirmation per week."
Set thresholds at the criterion level, not at the total score level. A rep with a 78% overall score but a 45% compliance score on a legally significant criterion needs a targeted PIP for that criterion, not an overall performance review.
Step 6: Measure Progress at 30, 60, and 90 Days
Performance management timelines that wait for a 90-day final review miss the feedback loops that allow mid-course correction. A 30-60-90 cadence produces three measurement points: an early signal at 30 days (is the trend moving at all?), a mid-point assessment at 60 days (is the pace sufficient to reach target?), and a final evaluation at 90 days (was the target reached?).
At the 30-day check, look for directional movement, not threshold achievement. A criterion moving from 45% to 52% in 30 days may reach 75% by day 90. A criterion that does not move at all in 30 days despite two coaching interventions requires a change in approach at day 30, not day 90.
Document all three review points in the employee record. The 30 and 60-day reviews should trigger either a continuation of the current plan or a documented modification.
| Measurement Point | What to Check | Decision |
|---|---|---|
| 30 days | Directional trend | Change approach if no movement |
| 60 days | Pace of improvement | Adjust intensity if pace is insufficient |
| 90 days | Threshold achievement | Continue, close, or escalate |
FAQ
How do you use feedback analysis to identify performance management issues?
Start with criterion-level QA data segmented by role and team. Flag criteria where performance falls below threshold after training was delivered: that combination indicates a performance management issue rather than a training gap. Aggregate feedback across call QA, manager observation, and self-assessment to build a complete picture. Measure behavior change at 30, 60, and 90 days against specific criterion targets.
What is the difference between a training gap and a performance management issue?
A training gap exists when a behavior falls below standard and no clear instruction on that behavior was delivered and retained. A performance management issue exists when instruction was delivered, understanding was confirmed, and the behavior still does not meet standard over a sustained period, typically 60 or more days. The intervention differs: training gaps require better instruction design; performance management issues require accountability structures and formal improvement plans. Insight7's criterion-level analytics help L&D managers make this distinction using call data rather than manager impression alone.
For HR and L&D teams building this diagnostic capability for the first time, start with a single high-impact criterion already scored consistently in your QA system. Map the before-and-after performance for that criterion over one quarter to establish your baseline methodology. The SQM Group's contact center quality benchmarks provide useful reference points for setting initial criterion thresholds in customer service environments.
See how feedback analysis connects to team-level coaching: explore Insight7's training tools.
