Assessment · Evaluation Design

Skills Gap Analysis — Sales Team

A Kirkpatrick Level 2 & 3 evaluation plan built to surface real performance gaps — not just quiz scores. Paired with scenario-based assessment items that mirror actual sales conversations and tracked via Tableau dashboards.

Tools: Excel, SurveyMonkey, Tableau Format: Mixed-method evaluation Audience: 47-person sales org

Project Snapshot

Problem

Sales training was measured almost exclusively by completion rates and post-quiz scores. Leadership had no visibility into whether reps could actually apply new skills in live deals — and conversion rates reflected it.

Design Goal

Build an evaluation architecture that distinguishes learning (L2) from behavior change (L3), using scenario-based items that require application — not recognition — and connect training outputs to business data.

Outcome

22-point increase in scenario assessment scores post-training. Manager observation data confirmed 78% of reps applied target behaviors on the floor within 45 days. Tableau dashboard now part of the quarterly L&D review.

Evaluation Framework

The plan targets two Kirkpatrick levels. Level 1 (satisfaction) was already tracked — the gap was in knowing whether training actually moved the needle.

Reaction

Already tracked via end-of-course satisfaction surveys. Not in scope for this project.

Learning

Did reps acquire the target knowledge and skills? Measured via scenario-based pre/post assessment items scored against a rubric.

SurveyMonkey scenario assessments (pre & post)
Excel scoring rubric with weighted criteria
Cohort-level gain score analysis

Behavior

Are reps applying new skills on the floor 30–60 days post-training? Measured via structured manager observation and CRM data pulls.

Manager observation checklist (standardized)
CRM activity data (30 / 45 / 60-day checkpoints)
Tableau dashboard for ongoing monitoring

Results

Business impact (quota attainment, win rate). Planned for Phase 2 once L3 baselines are established.

Scenario-Based Assessment — Sample Item

These items move beyond recall. Each scenario presents a realistic sales moment and requires the learner to choose the most effective response — scored against a four-point rubric.

Assessment Item — Level 2 · Scenario

Handling the "We Already Have a Vendor" Objection

—

SITUATION

You're 12 minutes into a discovery call with a mid-market prospect. The conversation has been going well — they've confirmed budget authority and acknowledged a pain point. Then the contact says:

"Look, we appreciate your time, but we've been with our current vendor for six years. Switching is a huge lift for our team and honestly not something we're looking at right now."

What is the most effective next move?

A Thank them for their honesty and offer to send a comparison one-pager they can review at their own pace — keeping the door open without pushing.

B Acknowledge the switching concern directly, then ask a focused question to understand what "not looking right now" is really about — is it timing, risk, or a genuine priority conflict?

C Pivot immediately to ROI data — explain that most customers recover switching costs within 90 days and share a case study from a similar company.

D Agree that switching is difficult and ask if there's a point-person on their team you should loop in for a more formal evaluation process.

Design Decisions

Scenario over recall

Multiple-choice recall items inflate scores without predicting behavior. Every assessment item places the learner inside a real sales conversation — they must choose a response, not a definition.

Weighted rubric scoring

Rather than binary correct/wrong, a four-point rubric captures degree of application. This produces richer data for cohort analysis and flags learners who are close — not just those who failed.

Pre/post design for gain scores

Administering the same scenario set before and after training isolates the effect of the intervention. Gain scores prevent high prior-knowledge learners from skewing overall results and expose where training added the least lift.

L3 tied to observable behavior, not self-report

Manager observation checklists — not self-assessments — anchor the L3 measurement. Calibration sessions ensured inter-rater reliability across 11 managers before data collection began.

Results

+22 pts

Average scenario assessment gain score (pre → post)

78%

Reps applying target behaviors at 45-day observation

Learners measured across 4 regional cohorts

Managers calibrated for L3 observation reliability

Want to see the full evaluation plan?

The complete package — rubric, SurveyMonkey item bank, observation checklist, and Tableau dashboard mockup — is available on request.

Get in Touch Back to Portfolio