The intranet page was published. The responses did not arrive

Imagine this internal request. The intranet article is live, the email has gone out and managers have been asked to remind their teams. From a communications perspective, the work can look complete. The team waiting for the form submissions still does not have the information it needs, so the process the communication was meant to support is still waiting.

My view is that an action-led internal communication ends when usable input reaches the person or team that needs it. The page is one component of that service. If you judge success at publication, you can report a completed communication while colleagues elsewhere absorb the failure through missing evidence, chasing, rework or delay.

That does not mean a low response proves the article was unread or badly written. It means you need to trace the whole request before deciding what to fix:

Start with the input the organisation needs

Suppose an organisation asks staff to complete a declaration, book a required appointment or provide evidence for a decision. Before anyone commissions an article, four things should be explicit:

• what input the organisation needs
• what makes that input usable
• who needs to receive it
• when they need it

Only then can you identify which staff need to act and what they need to understand.

This reverses a page-first production sequence. The request may arrive as “we need an intranet article and an all-staff email”. Those are proposed outputs. They say nothing about whether a page and email can produce the required response. A request for content is evidence of a possible need, not a settled solution.

I would not accept “awareness” as the objective when the organisation needs an observable action. Awareness may contribute, but it cannot supply a completed declaration, booking or form. The Government Communication Service makes a similar distinction in its Evaluation Cycle: producing content, reaching an audience, affecting understanding, changing behaviour and contributing to an organisational result are different levels of evidence.

The organisational result may depend on much more than communication, so this is not a claim that content caused the final outcome. It is a way to make the dependency visible. If a decision needs 200 valid responses by Friday, that requirement should shape the audience, route, page, form and follow-up. “Article published on Monday” is not an adequate substitute.

The broader measurement model belongs in the article on what internal communication can and cannot do. The important point here is narrower: the operational endpoint changes what you inspect.

Discovery is part of the service

Publishing a page makes it available. It does not establish that the staff who need it encountered it.

“All staff” does not answer “who must act?” Different roles, locations, working patterns and levels of access can change both the relevance of a request and the route by which it reaches people. If the audience itself is unclear, start with a map based on what each group must do.

Then name the discovery route for each group. It might be a targeted email, an intranet link, a task in a staff system, a team meeting or a manager briefing. If the plan relies on managers directing their staff to the page, that is a real step in the service. It needs an owner, timing, a clear instruction and a way to tell whether the relay happened.

Manager involvement is not inherently a defect. Managers can add context and reach colleagues missed by a central digital channel. The risk is treating the relay as automatic. Local Government Association guidance on communicating through line managers recognises both their value and the burden placed on busy managers, and recommends notice, coordination, support and a clear request.

This is where diagnosis matters. A person may not act because they never encountered the request. They may have encountered it and decided it did not apply to them. They may intend to act but lack time, access, authority or the evidence the form requires. Treating all of these as an “engagement” problem hides the decision you need to make.

One study of bulk email in a large university found that recipients could open organisation-wide messages without reading them in detail or acting on them. It is a useful warning against equating an open with attention, not a benchmark for other organisations. The authors recommend measures tied to the complete transaction when action is required.

Give the page one precise job

Once staff reach the article, the content has a specific job. A person should be able to tell:

• why the request applies to them
• what they must provide or do
• when it is due
• what will happen after they act
• where to get help

The email, page and form must agree on eligibility, timing and required evidence. Conflicting instructions can stop the action because the person has to decide which source to trust.

This is where content design earns its place. Start with what the person needs to do and treat the requested format as a proposal. The fuller content-design method is covered separately. Here, the content test is whether the page carries the person from discovery to a possible, informed action.

Clear prose serves people who reach a relevant request and have a realistic opportunity to act. A missed audience, impossible deadline or unusable form needs a different remedy. The GCS behavioural approach uses the COM-B model to distinguish capability, opportunity and motivation. That is a useful guard against blaming the message, or the employee, for every incomplete action.

My test is simple: before rewriting the page, identify what evidence would show that the problem is actually in the page.

Follow the response into the receiving team

The journey does not necessarily end when someone presses submit.

Check whether the action is findable from the article, whether the form works with the devices and access available to staff, and whether the amount of information requested is proportionate. Then follow a completed response through its confirmation, validation and transfer to the right queue. Record whether the requesting team can use it and how incomplete or duplicate responses are handled.

A completion count can look healthy while the operational input remains poor. Required fields may be missing. Attachments may be inaccessible. Responses may sit in an inbox nobody owns. These are possibilities to investigate, not claims about every internal form.

This boundary matters for content designers. Content can expose a process or system defect and route it to the person who owns the remedy. It cannot make incompatible systems exchange data or give a team authority it does not have. Calling every break a content problem creates another false green: the words improve while the service remains stuck.

Diagnose the chain before changing the page

Use the smallest record that makes the handoffs visible.

Stage	Question	Evidence to inspect	Possible owner
Intended staff	Which people must act for the required input to exist?	Required roles or locations, staff denominator, operational requirement	Requesting team with internal communications
Discovery	How should those people encounter the request?	Targeted delivery, intranet routes, referrals, manager confirmation	Internal communications, channel or manager owner
Understanding	Do they know why it applies, what to provide and when?	Short comprehension checks, questions and support contacts	Content owner
Action	Can they complete the task in their working context?	Starts, completions, errors, accessibility findings and observed use	Form, process or system owner
Receipt	Did usable input reach the team that needs it?	Accepted submissions, missing fields, timestamps, chasing and rework	Requesting or operational team

The first weak measure is a place to investigate. It is not proof of the root cause. Low form completion does not establish a discovery failure. High page views do not establish that the right staff understood the request. A successful submission does not establish that usable data reached the next handler.

My practical rules are these:

• define the staff action and operational recipient before producing the page
• name manager referral as a handoff when the journey depends on it
• inspect the form and receiving process with the content
• do not substitute publication, views or opens for the required action
• do not infer why staff failed to act from the completion rate alone

Without that shared evidence, each team can complete the component it owns while the missing work lands between them.

Take one current internal request. Write down the intended staff, discovery route, required understanding, action and operational recipient. Beside every transition, record the evidence that would show it happened. The first blank cell is where the next investigation starts.