11.2 Barriers to Communication and Cognitive Biases
Key Takeaways
- Barriers to communication include organisational hierarchy, organisational culture, technology, amount of information, type of information, and attitudes and perceptions
- Cognitive biases are filters people use when receiving communication
- Confirmation bias is confirming one's own existing beliefs
- Status quo bias is the preference for keeping things as they are
- Availability bias focuses on more memorable or easily available information, and loss aversion means losses are felt more strongly than gains
Two distinct problems interfere with change communication: barriers that obstruct the message and biases that distort how it is interpreted.
Barriers to communication
CM3 names six:
| Barrier | How it obstructs |
|---|---|
| Organisational hierarchy | Messages passing through layers are filtered, softened or lost; upward messages especially so |
| Organisational culture | A culture where challenge is unwelcome suppresses the feedback the change team needs |
| Technology | Systems people cannot access, formats they cannot open, or channels not everyone uses |
| Amount of information | Too much and it is not read; too little and rumour fills the gap |
| Type of information | Content mismatched to the audience — technical detail to a general audience, or vague reassurance to specialists |
| Attitudes, perceptions | What the receiver already believes about the change, the sender and the organisation |
Hierarchy and culture are the structural ones
Hierarchy degrades messages in both directions. Downward, each layer adds its own interpretation and omits what it considers unimportant, so the message reaching the front line can differ materially from the one sent. Upward, bad news attenuates at every level, which is why sponsors are often the last to learn a change is failing.
This is precisely the problem change agent networks solve. A network provides a route that bypasses the hierarchy in both directions — trusted, 'one of us' messages down, and unfiltered reality up.
Amount and type of information
These two are the barriers change teams create themselves. Amount is the more common failure: a change programme generating weekly bulletins, monthly newsletters, town halls and intranet updates produces a volume that guarantees none of it is absorbed. Fewer, better-targeted communications outperform more.
Type is a segmentation failure. A single message written to satisfy every audience satisfies none — it is too detailed for the general population and too vague for the people who need specifics.
Cognitive biases
Behavioural psychologists have identified many cognitive biases (Kahneman, 2012). These are the filters people use when receiving communication. CM3 examines four:
| Bias | Definition | How it shows up in change |
|---|---|---|
| Confirmation bias | Confirming own beliefs | Someone who has decided the change will fail reads every setback as proof and discounts every success |
| Status quo bias | Keeping things as they are | "Why do we have to change everything? There are many systems that are still effective" |
| Availability bias | Focus on more memorable or easily available information | One vivid failure story from a colleague outweighs statistics from twenty successful sites |
| Loss aversion | Impact of losses felt more strongly than gains | A change offering a clear net benefit is resisted because the specific losses are more salient than the diffuse gains |
Distinguishing status quo bias from loss aversion
These two are easily confused because both produce resistance to change. The distinction:
- Status quo bias is a preference for things staying as they are, expressed as a defence of the current state. "What we have works."
- Loss aversion is about the asymmetric weighting of losses against gains. The person may accept the change is better overall but still resist, because what they lose feels bigger than what they gain.
A statement like "Why do we have to change everything? There are many systems that are still effective" is a defence of the current state — status quo bias.
Working with biases rather than against them
Biases cannot be argued away — attempting to correct someone's confirmation bias with more evidence typically strengthens it. What works is designing communication that accounts for them:
For loss aversion, name the losses explicitly and specifically before describing the gains. Unnamed losses loom larger than named ones, and acknowledging them removes the sense that the change team is hiding something. This is the same discipline as the dis-benefit conversation and the change-curve guidance not to minimise losses.
For status quo bias, do not attack the current state. Telling people that what they have built is deficient triggers defence. Appreciative Inquiry's approach — identifying what works well and building on it — is designed for exactly this, and it explains why AI increases motivation to change.
For availability bias, provide vivid, concrete, memorable positive examples rather than aggregate statistics. One credible story from a peer site who has already made the change beats a percentage.
For confirmation bias, the strongest counter is having the message carried by someone the receiver already trusts — a change agent from their own area — because the source is the thing their filter is applied to.
Barriers and biases together
A useful way to hold the two apart: barriers stop the message arriving; biases change what it means when it does. A message blocked by hierarchy never reaches the receiver. A message filtered by confirmation bias arrives fully and is interpreted as confirming what the receiver already thought.
A department head argues against a platform migration on the grounds that the existing tools work perfectly well and nothing about them needs replacing. Which cognitive bias does this argument most clearly display?
Which of the following is one of the six barriers to communication CM3 identifies?
A team dismisses evidence from twenty successful implementations because one colleague described a bad experience at another site in vivid detail. Which bias is operating?