Internal communications and employee engagement: what comms can and cannot do

What communication can and cannot do

Employee engagement is mostly made of things communication does not control: the work itself, the quality of management, and the terms of employment. Internal communication is the layer underneath. It decides whether people know what to do, whether they know what is going on, and whether they can be heard.

That is a real contribution, and it has a limit worth stating plainly. Good internal communication cannot make a badly run organisation engaging. Bad internal communication can damage an otherwise well-run one, because silence reads as disregard and spin reads as contempt. The causal direction communication owns runs through failure: it is easier to lose engagement by getting this wrong than to manufacture it by getting it right.

How communication builds trust

Trust in internal communication accrues the same way it does anywhere: through kept signals. The 3 that matter are telling people things before they had to ask, saying what is not yet known, and doing what was said after the meeting.

Accuracy is the minimum. A message can be accurate and still spend trust by arriving after the question it answers, or by answering a question nobody asked while ignoring one everybody did.

The routes through internal communications

The pages around this one each carry a single job:

Your problem	The page
You do not know who you are talking to	How to map your internal communication audiences
The message also goes outside	Internal versus external communication, and how to translate between them
Something has gone wrong and timing is measured in minutes	Crisis communication: what changes, and what you can prepare
The whole practice needs its framework	What a stakeholder engagement framework is

What leadership's part consists of

Leaders are the senders people believe. A communications function can distribute decisions, but when every message arrives from the function, the function becomes the face of decisions it did not make. The leader's part is 3 behaviours: send the messages that matter personally, answer what the reply route brings back, and be seen changing something because of feedback. Anything less delegates the appearance of listening without its substance.

Closing the loop is the part that gets dropped

Asking people what they think is easy to schedule and easy to be praised for. Telling them what happened to their answer is neither, and it is the half that does the work.

The Teal Book, the UK government's project delivery standard, states the mechanism plainly: where someone takes the time to contribute their views and then hears nothing further, they can reasonably conclude their input was not valued or has been ignored. Its remedy is specific, and more demanding than most feedback processes attempt. Report back not only on what was decided, but on why.

That second half matters because it is the only version that survives disagreement. "We considered your suggestion and decided otherwise, for this reason" leaves a person informed and still willing to contribute. Silence leaves them with 2 available conclusions, that nobody read it or that nobody cared, and both of them are corrosive.

The Teal Book adds a related point about difficult messages: they land better delivered directly than discovered through another channel. Even an unwelcome decision, communicated in person by whoever made it, signals that the relationship is still worth the awkwardness. The same decision reaching someone through a forwarded email or a colleague's summary says the opposite, whatever the words say.

This is why the implemented feedback number matters more than the response rate. A consultation with 80% participation and no visible consequence teaches people that participating is pointless, and it teaches faster than a consultation nobody answered. The second round is always smaller, and the reason it is smaller is rarely recorded.

One caution belongs beside this. Closing the loop does not mean acting on everything, and a function that tries will grind to a halt and lose credibility from the opposite direction. What it means is that every input gets an outcome: adopted, declined with a reason, or deferred with a date. Three possible answers, all of them cheap, and any of them better than none.

During change

Change multiplies the questions communication has to answer, and multiplies the cost of silence. The machinery does not change. The audience map says who must act on the change. The translation worksheet separates the internal version from the external one. The holding structure, meaning what is known, what is not yet known, and when the next update comes, carries the period where answers do not exist yet.

One limit is worth naming. Communication cannot make an unwelcome change welcome. What it can do is keep trust intact while the change lands, by being early, honest about trade-offs, and specific about what is still open. People accept decisions they dislike more readily than decisions that arrive unexplained.

The minimum viable measurement plan

This is the internal adaptation of the 5 numbers in measuring stakeholder communication. Each number needs a baseline, a threshold and a review date, or it is a statistic rather than a plan.

Number	What it asks	What it catches
Coverage	Did each row of the audience map receive what it needed, on the channel the map names	Rows the map misses: groups learning things late or secondhand
Comprehension	Can someone in the group say what changed and what they do next	Messages that were sent and not understood
Cadence held	Did the planned rhythm of communication actually happen	Plans that decayed into silence
Implemented feedback	How many replies changed a decision or a message	Listening that was collection
Voice	Do people bring questions to the named reply route before taking them elsewhere	Whether the trust argument is working

These numbers replace the activity metrics this field tends to reach for: open rates, click-throughs, attendance counts, sentiment scores. Those measure distribution, and distribution is not communication. An email can be opened and useless. A meeting can be attended and misunderstood. Coverage and comprehension catch what activity counting cannot.

Test comprehension on the next announcement

Take the next internal announcement and name the 3 groups who must act on it. Send to the map, not the all-staff list. One week later, ask 5 people across those groups two questions: what changed, and what do you do next. Full answers mean communication happened. Anything else means content was distributed, and the number to check first is comprehension.