A Practical Guide to Three-Phase Assessment Design for Leadership Development

A friend of mine who runs leadership development programmes struggled with renewals for three years. Client feedback was consistently positive—they'd say things like "really valuable"—yet when it came to renewal conversations, there was always hesitation. Then he did one thing differently: he embedded three assessment points into the programme, and renewal rates jumped that very year.

I asked him: what was the secret? He said: no secret, really. I just made it possible for clients to see their own change.

That's the biggest blind spot in many coaching programmes: the impact is real, but invisible. Learners say they've "gained a lot", but how much? In which dimensions? How does it map back to the original objectives? Without clarity on these questions, procurement teams struggle to justify renewal.

Understanding Kirkpatrick's Four-Level Model

To design effective assessment, you first need to know what you're measuring. The Kirkpatrick Model has been the gold standard in training evaluation for decades, breaking assessment down into four levels:

Most training programmes only ever reach Level 1—the satisfaction survey. Not because Levels 2, 3, and 4 don't matter, but because they're more effort. Yet that effort is exactly what proves value.

The good news: you don't need all four levels. Nail Level 2 and Level 3, and you'll already have a compelling case.

The Three-Phase Assessment Framework

The most effective leadership programmes I've worked with typically have three assessment touchpoints. They're not "add-on homework"—they're part of the programme experience itself:

Designing the Baseline Assessment

The baseline is the foundation of the entire system, and where you should invest the most thought. A few principles:

Focus on core programme dimensions—don't overdo it

For a typical 6-month leadership programme, 4-6 core dimensions are sufficient. 3-4 questions per dimension, keeping the total under 20 questions. Beyond that, completion rates drop sharply—and more questions don't necessarily mean better data.

Common leadership dimension references: self-awareness & emotional regulation, decision quality & risk judgment, influence & communication style, team building & delegation, strategic thinking & prioritisation. Which ones you choose depends on your programme positioning.

Use Likert scales, not open-ended questions

The baseline's purpose is to generate quantifiable data for before/after comparison. 5-point or 7-point scales work best. For example: "I can maintain clear decision-making under pressure"—from 1 (strongly disagree) to 5 (strongly agree). These are straightforward to write and easy to analyse.

Questions must be directional, not vague

Good questions have directionality—you can immediately tell what a "high score" means. Avoid questions like "I am satisfied with my leadership"—that's dimensionless; neither high nor low scores tell you anything useful.

The Logic Behind Mid-Programme Checks

The mid-programme check is the most commonly overlooked element, yet it often has the biggest impact on sustained learner engagement.

It doesn't need to be complex. I typically recommend 6-8 questions, structured in two parts: first, re-ask 4-5 "indicator questions" from the baseline (for data comparison); second, add 2-3 behavioural questions like "Over the past 6 weeks, which new communication approaches have you applied?"—these don't need quantification, but they help learners articulate their own growth.

Once the mid-check is complete, send learners a simple progress update the same day. Even something as basic as "your emotional regulation score has improved by 0.6 points since baseline" creates a powerful motivational effect through immediate feedback.

The Completion Report Is Your Most Important Deliverable

The completion assessment uses the exact same questions as the baseline, but this time it comes with a data-backed report. This report should include:

• Before/after comparison by dimension: use charts to show score changes across each dimension from baseline to completion
• Areas of strongest improvement: which dimensions shifted the most—this is what learners want to see
• Areas still to develop: where there's still room to grow—honest data is more credible than "all excellent"
• Evidence of behavioural change: pull in behavioural descriptions learners provided during the mid-check

This report has two audiences: the learner themselves, and the HR or leadership team who purchased the programme. For the learner, it's a record of growth. For the buyer, it's proof of programme effectiveness.

Common Design Pitfalls

Pitfall 1: Treating assessment as "homework" tacked onto the end. If learners see assessment as a form they're forced to fill out near the end, both completion rates and data quality will suffer. Position assessment as an integral part of the programme from day one, explaining its purpose clearly.

Pitfall 2: Using different questions for baseline and completion. I've seen people switch to "better" questions at completion—making the data completely incomparable. Baseline and completion must use the same core question set.

Pitfall 3: Testing knowledge instead of behaviour. Questions like "Do you understand the Situational Leadership model?" have little value. Ask behaviour instead: "When a direct report shows inconsistent performance, do you adjust your management approach?" Behavioural data is what truly persuades.