Unconscious Bias Coaching - Does It Really Work?

Bias rarely shows up as an obvious bad intention; it usually slips into small decisions about who gets heard, who gets trusted, and who gets promoted. In practice, unconscious bias coaching works best when it is tied to real decisions, not abstract theory. This article explains what that kind of coaching changes, where it helps, where it falls short, and how I would evaluate it in a UK workplace.

What this coaching actually changes

I use this kind of coaching to mean a structured process that helps a person notice where bias shows up, slow down automatic judgments, and build better decision habits. That matters most for managers, recruiters, project leads, and senior specialists whose choices affect access to opportunity.

Acas describes unconscious bias as a tendency to make assumptions that are not right or reasonable, which is why it can quietly shape decisions about hiring, promotion, performance, or even who gets interrupted in a meeting. The point of coaching is not to shame people for having a brain that makes shortcuts; it is to help them catch those shortcuts before they harden into patterns.

• Hiring - coaching can surface “fit” language that really means similarity bias.
• Feedback - it can reduce vague criticism that lands harder on some people than others.
• Stretch assignments - it can make managers notice who is repeatedly overlooked for visible work.
• Meetings - it can help leaders track who speaks, who gets interrupted, and whose ideas get repeated by others.
• Promotion decisions - it can force a clearer link between evidence and readiness.

That is the core value: the work moves from general awareness to specific decision points, which is where bias usually does its damage. From there, the real question is why awareness alone so often stalls.

Why awareness alone rarely changes behaviour

The idea that bias can be “solved” by a single workshop is still widespread, and it is still too optimistic. The EHRC review found a mixed picture: awareness can improve, more sophisticated interventions can reduce implicit bias for a period of time, and behaviour-change evidence is limited. That is the gap I care about, because people can leave a session convinced and still behave exactly the same under pressure.

There are three reasons this happens again and again.

• Habits beat insight - when deadlines are tight, people fall back on familiar shortcuts.
• Good intentions are not a system - if the process is vague, bias has room to reappear.
• Self-awareness is uneven - many people can explain bias in theory without spotting it in themselves.

That is why I never treat awareness as the finish line. Awareness is useful, but it becomes valuable only when it leads to better prompts, better structure, and better accountability. Once that is clear, the design of the programme matters much more than the label on it.

What a useful programme looks like

Real cases beat generic examples every time. If someone is coaching a hiring manager, I want the discussion to be about the last shortlist, the actual criteria used, and the moments where “gut feel” overrode evidence. That makes the work concrete enough to change behaviour.

That structure also fits UK workplaces well, because the best programmes here tend to be framed as better leadership and fairer decision-making, not as a compliance exercise. From there, the next step is deciding how the sessions themselves should work.

How I would structure the sessions

I prefer a simple sequence that moves from reflection to action. The goal is not to over-engineer the process; the goal is to make it hard for bias to hide inside routine decisions.

1. Map the pattern - ask the participant to recall a recent decision and identify what was noticed first, what was ignored, and what felt “obvious.”
2. Name the trigger - explore the likely bias pattern, such as affinity bias, confirmation bias, or halo effect. Affinity bias means favouring people who feel similar; confirmation bias means looking for evidence that supports an early view; halo effect means letting one positive trait colour the whole judgment.
3. Choose one behaviour - the change should be visible, such as using the same shortlist criteria every time or asking for counter-evidence before a promotion decision.
4. Rehearse the replacement action - use scripts, checklists, or a short role-play so the new behaviour is easier to repeat under pressure.
5. Review the evidence - compare what happened in real meetings or decisions over the next few weeks, not just what people intended to do.

For a manager, the practical shift might be tiny but powerful: instead of saying, “She feels like a good fit,” the manager is coached to say, “What evidence did we collect against the agreed criteria?” That small change is often where fairness becomes measurable.

This is also where group coaching and one-to-one coaching diverge. Group work can create shared language and peer accountability, but only if people feel safe enough to discuss actual decisions rather than perform the right answers. One-to-one work goes deeper into personal habits, but it needs a real connection to the participant’s role.

Coaching works best when the system changes too

If I had to be blunt, I would say coaching is a decision-quality tool, not a complete inclusion strategy. It helps individuals interrupt bias, but the surrounding process still decides how much damage bias can do.

In practice, I want all three levers working together. Structured interviews, standard scoring, blind sifting where appropriate, and calibration sessions do more than remind people to be fair; they make it harder to drift into vague, biased judgment. The EHRC review also points to this wider-programme logic, noting that awareness alone is not enough for long-term organisational change.

The most useful sequence is usually simple: coach the decision-makers, improve the process, then revisit the same decision point after a cycle or two. That leads directly to the question leaders ask next, which is whether any of this is actually working.

How to tell if it is working

I would never rely on “people liked the session” as my main success measure. Satisfaction matters a little, but it tells me more about the room than about the workplace.

Instead, I look for changes in behaviour and decisions over time. A basic review after one quarter and again after half a year is often enough to see whether the coaching changed anything meaningful.

• Decision consistency - are interview scores, promotion criteria, and feedback standards being used more consistently?
• Process adoption - are managers actually using the checklist, rubric, or structured questions they agreed to use?
• Participation patterns - are quieter voices getting more space in meetings and project allocation conversations?
• Outcome gaps - are there signs that progression, ratings, or stretch opportunities are becoming less skewed?
• Quality of reflection - can participants name their own bias triggers and describe what they did differently?

What I do not want is a programme that claims victory because people scored well on a survey. Better awareness is useful, but if the same teams still make the same choices, then the intervention has not done enough. That is why the common mistakes matter so much.

Common mistakes that undermine the work

Most disappointing programmes fail for predictable reasons. The idea is not bad; the execution is too shallow, too generic, or too isolated from the work itself.

• Making it a tick-box - one mandatory session with no follow-up rarely shifts anything durable.
• Talking about bias in the abstract - people need the discussion anchored in their own hiring, feedback, or promotion decisions.
• Ignoring the process - if criteria are vague, coaching has to do too much heavy lifting.
• Skipping manager buy-in - if leaders do not reinforce the new behaviour, the habit disappears fast.
• Measuring only attitudes - confidence is not the same as changed conduct.
• Turning coaching into blame - defensiveness is a poor learning environment and usually a sign the framing is wrong.

I also see programmes stumble when they assume bias is only an individual flaw. Sometimes the real problem is the design of the shortlist, the structure of the meeting, or the way performance evidence is gathered. If coaching does not touch those points, it will be asked to compensate for a broken system.

That is why the next move should be practical, narrow, and visible.

Where I would start in a UK team

If I were starting from scratch in a UK organisation, I would choose one decision point with real consequences, such as hiring, promotion, or stretch assignment allocation. I would then coach the people involved in that decision, not as a standalone exercise, but alongside one small process change that makes fairer behaviour easier to repeat.

1. Pick one high-stakes decision where bias is likely to matter.
2. Use coaching with the managers or recruiters who make that decision.
3. Add one guardrail, such as structured criteria or a short review checklist.
4. Agree on two or three metrics to review after the next cycle.
5. Repeat the same pattern before expanding to another team or function.

That is the version of the work I trust most. When unconscious bias coaching is embedded like this, it stops being a slogan and starts changing how people hire, review, delegate, and promote.