Behavioural Change Toolkit · Diagnose

Barrier Analysis — COM-B Interview Guide

Six to ten conversations of 30 minutes will tell you more than any survey. This is the guide.

---

Ground rules

1. Talk to people who do the work. Not their managers describing what they think happens.
2. Six is usually enough to hit saturation on barriers. Ten if roles vary widely.
3. Never sell. The moment you defend the change, the data stops. You are there to learn, not to persuade.
4. Ask about last time, not in general. “Walk me through the last time you…” gets you reality. “How do you usually…” gets you the official version.
5. Confidentiality, stated plainly, and honoured. “Nothing you say goes back with your name on it.” Then don’t.
6. Watch, if you can. Twenty minutes sitting beside someone doing the task beats an hour of description. You will see workarounds they’ve stopped noticing.

---

Opening (2 min)

“I’m trying to understand how the work actually happens, not how it’s supposed to happen. Nothing here goes back with your name attached, and there’s no answer that gets anyone in trouble. If the honest answer is ‘we ignore that step’, that’s the most useful thing you can tell me.”

---

The core sequence

Current state

• Walk me through the last time you [did the relevant task]. Start from the beginning.
• What did you actually do — including anything that isn’t in the official process?
• Where does it get annoying?
• If you had to train someone new tomorrow, what would you tell them that isn’t written down anywhere?

(That last question is gold. It surfaces the real process.)

Capability

• Has anyone shown you how to do the new way? What was that like?
• Is there any bit where you’re not sure what you’re supposed to do?
• Are there parts where you’d have to ask someone?

Opportunity — physical

• How long does the new way take compared to the old?
• Do you have everything you need — access, kit, licence, the right screen?
• When in your day is this supposed to happen? What else is happening then?
• What would you have to stop doing to fit this in?

Opportunity — social

• What do the people around you do?
• What does your manager actually ask you for?
• If you did it the new way and nobody else did, what would that be like?
• Who around here do people listen to on this stuff?

Motivation — reflective

• What do you make of the change, honestly?
• Who do you think it’s for?
• What would have to be true for you to think it was a good idea?
• What’s the argument against it that nobody’s said out loud?

Motivation — automatic

• When you’re busy or under pressure, what happens?
• Do you find yourself doing the old thing without thinking?
• What triggers you to remember?

Closing

• If you could change one thing about how this is being done to you, what would it be?
• What am I not asking about that I should be?
• Who else should I talk to?

---

The three highest-yield questions

If you only get ten minutes:

1. “Walk me through the last time you did this.”
2. “What would you have to stop doing to fit this in?”
3. “What’s the argument against this that nobody’s saying out loud?”

Question 3 gives people permission to voice the thing everyone knows and nobody will put in a survey. It is frequently the single most valuable sentence in your whole diagnosis.

---

Capturing it

Score each conversation immediately after — 1 (total blocker) to 5 (no issue):

Person / role	PhysCap	PsyCap	PhysOpp	SocOpp	RefMot	AutoMot	Killer quote
Mean

The lowest mean column is your primary barrier. Take it to Intervention Design Worksheet.

---

Presenting this to leaders

Three things, in this order:

1. The scores — one chart, six bars. Concrete, non-emotive, hard to argue with.
2. Three verbatim quotes, anonymised. Quotes do the persuading that data can’t. Choose the ones that are uncomfortable but fair.
3. One thing the organisation has done to itself that’s blocking the behaviour. There always is one, and finding it is what makes you useful rather than just critical.

---

→ COM-B and the Behaviour Change Wheel · Resistance Decoder · Behavioural Diagnosis Canvas

← Back to the Behavioural Change Toolkit