20 'killer' lessons for securing successful organisational change
Transformations don’t usually collapse at launch. They decay after the applause, when old habits are still easier than new ones. Here’s what really causes that failure — and how to prevent it.
This Thursday, I co-host a webinar with my good friend Richard Jordan, former MD at Barclays for communications. We will unpick the challenge many in business face - how to make change ‘stick’ and why is it that so many organisational transformations fail to achieve their stated goals.
Before that webinar, I have sought to outline some of the challenges, with an overview of what I have experienced over the last 30 years of working and advising on organisational change. What works - and what doesn’t.
First, let’s be clear. Organisational transformations are being implemented everywhere - new strategies, new structures, new ways of working.
And yet, far too many fail to deliver what they promise. Not just in headline results, but in everyday reality: behaviours don’t change, decisions don’t speed up, and the organisation quietly drifts back to what it was doing before.
This isn’t because leaders are stupid or employees are stubborn. I am absolutely of the view that it is because most transformations are designed around ideas (vision, culture, communication) rather than around behaviour, systems, and constraint.
In practice, organisations are very good at launching change and very bad at securing it.
The failure mode is usually subtle. The programme looks busy. Dashboards glow green. Engagement scores hold up. But underneath, people continue to prioritise the old metrics, use the old tools, and work around the new processes. Compliance replaces commitment. Adoption becomes theatre.
So why do transformation programmes keep failing? And type of signals should you look for when real change has happened in your organisation?
Here are 12 reasons why:
1. You’re selling belief when you should be redesigning behaviour. Most programmes try to “win hearts and minds” first. But behaviour follows friction, incentives, and convenience. That’s why the “failure rate” of so many change programmes I’ve observed has been so stubbornly high.
Positive change signal: the old way becomes harder (retired tools, closed loopholes), not just less fashionable.
2. The old system is still alive so people rationally revert. If you run legacy and new processes in parallel, you’re not transforming - you’re adding choice, and choice defaults to habit under pressure. BCG’s research on digital transformation highlights that success odds can move dramatically when specific enabling conditions are in place (their “30% to 80%” framing when key factors are done well).
Positive change signal: reversal is inconvenient (legacy approvals removed, duplicate reporting killed, governance simplified).
3. Leaders don’t realise they are the real policy. Employees watch leaders to infer the actual rules. If leaders bypass the new process “to get it done,” they quietly revoke the change. This is why sponsor effectiveness is so predictive. Over the last 20 years, I’ve seen that change projects with extremely effective sponsors are far more likely to meet objectives than those with ineffective sponsorship (their published comparisons show large gaps).
Positive change signal: leaders take the pain publicly (no exceptions for “special cases”; consistent reinforcement of non-negotiables).
4. You confuse engagement with adoption (and green dashboards with reality). A transformation can be “well received” and still fail - because the real question is behavioural: what decisions are made, what trade-offs are chosen, where time goes. Prosci, a change management consultancy, has repeatedly reported that successful transformations correlate strongly with change management effectiveness (e.g., projects with effective change management far more often meet/exceed objectives than poor change management).
Positive change signal: you can point to observable shifts (cycle times, decision rights used properly, fewer escalations, less duplication) not just sentiment.
5. You’ve created awareness but not sustained behaviour change. Workshops and comms create understanding. They don’t create durability. A common pattern is “applause then drift.” That’s exactly the failure mode I see far too often: leaders can start change; they struggle to prevent regression.
Positive change signal: after 90 days, the organisation is still doing the new thing without needing another comms push.
6. Cognitive load kills adoption (because you’ve added work, not removed it). Transformations fail quietly when they compete with day jobs and other initiatives. There’s a growing body of evidence on information overload increasing with digitalisation.
Working signal: the change subtracts workload (fewer reports, fewer approvals, fewer meetings) rather than adding “one more layer.”
7. Psychological safety isn’t “nice”; it’s an early-warning system. Unsafe organisations don’t surface problems until they’re expensive. Amy Edmondson’s foundational research links psychological safety to learning behaviours in teams - crucial during change because learning requires speaking up, admitting confusion, and reporting errors early.
Working signal: bad news travels fast (issues raised early; fewer surprises; teams admit what isn’t working without punishment).
8. You don’t use a real Theory of Change; you design activities, not outcomes. Most transformations start with a programme plan, not an outcome logic. A practical Theory of Change works backwards: outcome, then behaviours, then system enablers and, finally, interventions. Systems beat stories every time.
Working signal: you can name 3–5 behavioural signals that make success inevitable (and you measure those, not just outputs delivered).
9. The middle layer is overloaded and the change collapses in translation. Transformations often die in the “squeezed middle”: middle managers asked to execute strategy while absorbing churn, ambiguity, and employee emotion. HBR recently argued that traditional top-down transformation is producing weak lasting results and that middle managers are pivotal
Working signal: managers have authority plus capacity (decision rights clarified, trade-offs permitted, time freed, local constraints removed). If managers are only given comms packs, you’ve built a cascade—not a change engine.
10. You’re trying to transform while your governance makes decisions slower. A surprisingly common outcome: transformation adds “steering groups,” “design authorities,” and “alignment forums” and suddenly nothing moves. This is where project-performance data is a useful proxy. One organisation ‘ve advised reported significant waste due to poor project performance in its reporting (missed deadlines, scope creep, overspend), which is exactly what bloated governance produces.
Working signal: decision latency falls (fewer handoffs; fewer approvals; shorter cycle times), and teams can describe who decides what without guessing.
11. Informal networks decide whether change sticks (and you’re not measuring them). Real sentiment lives in peer conversations, not town halls. If you only measure “official” channels, you’ll miss early resistance and workarounds. Your stimulus doc calls this out directly: informal networks are the real sentiment barometer.
Working signal: employee research captures local truth (credible peer voices; rapid pulse checks; qualitative listening that identifies workarounds, not just opinions).
12. You never stop anything so everything decays into incoherence. People don’t resist change volume; they resist incoherence. When you keep stacking initiatives, you destroy attention and create change fatigue by design (again: your stimulus doc’s point - change fatigue is often a design failure).
Working signal: leadership can name what has been stopped, not just what has been launched.
Final thoughts…
Most transformations do not fail because people are resistant or leaders are uncommitted. They fail because the organisation’s underlying system continues to reward the old way of working. New language is introduced, new programmes launched, and new charts published - but the practical realities of how work gets done remain unchanged.
Under pressure, people do what is easiest, safest, and most recognised. When the system still points backwards, so does behaviour.
The uncomfortable truth is that sustainable transformation is not primarily a communication or culture problem. It is a design problem. It requires leaders to work backwards from outcomes to behaviours, and from behaviours to constraints, incentives, and decision rights. It requires stopping work, not just starting initiatives. And it requires consistency over time, long after the initial excitement has passed. Durable change is not created by belief alone; it is created when the new way of working becomes harder to undo than the old one.
The test of transformation, then, is not whether people understand it or support it. It is whether the organisation now makes different choices by default. When decision-making speeds up, when old habits become inconvenient, when leaders follow the same rules as everyone else, and when employees can feel work getting simpler rather than heavier, change has moved from announcement to reality.
At that point, transformation is no longer a programme. It has become how the organisation operates.
Register for this Thursday’s webinar to hear and take part in the discussion on why change fails. Click here.