In institutional communication, there is a very common blind spot: we make decisions based on intuition, experience and isolated signals… and then we are surprised when a campaign fails to take off, when a message is misinterpreted or when a rumour sticks around for weeks.
It’s not a lack of talent. It’s a lack of useful, carefully gathered information.
The good news is that you don’t need a huge survey to improve your decisions. You can do applied, light and ethical research that gives you actionable answers in days (or a few weeks). The key is twofold: ask the right questions and turn what you find into concrete decisions, not a report that stays in a drawer.
Although this is not a problem exclusive to institutions, in many companies, decisions are also made based on isolated signals. I focus on institutions because the demand for accountability and public impact makes going ‘blind’ more noticeable.
This blog is a practical guide for communication teams in institutions and organisations that want to investigate without invading or skewing the investigation, with small, realistic and replicable methods.
When someone asks you to ‘measure trust’, it is easy to imagine a huge questionnaire and a sample that never arrives. But in practice, the questions that best help you decide are often simpler and more observable.
Think of questions that are noticeable in behaviour, interaction or friction. Questions that, if answered, allow you to adjust the channel, tone, clarity or process.
For example:
What part of the message is not understood?
It’s not “whether they like it.” It’s whether they understand it. If people confuse a requirement, misinterpret an exception, or fail to distinguish between “recommended” and “mandatory,” your communication has a comprehension problem, not a creativity problem.
Where do people get stuck?
In procedures, calls for applications, appointments, assistance, and public consultations. Every abandonment, every repeated call, every email saying “I don’t know how to do it” is a piece of data. There is friction there, and friction needs to be investigated.
What is repeated as rumour or misunderstanding?
Rumours do not appear out of thin air. They tend to grow where there is a lack of information, ambiguity or prior mistrust. If you detect them early, you can decide whether to clarify, change the order of the information, activate spokespersons or reinforce part of the narrative.
What signs of trust (or mistrust) appear in the relationship with the institution?
Here, ‘trust’ is not a big word. These are small signs: people who believe that ‘this has already been decided’, that ‘there is no point in participating’, that ‘they are not listening’, or, conversely, people who recommend a channel because ‘it responds’. This can be investigated without asking Do you trust us?’ directly.
Which groups are left out without you seeing it?
Accessibility and equity are not values in a paragraph. They are decisions: schedules, formats, languages, easy reading, and digital barriers. Investigating inclusion often means detecting who is not reaching and why.
In short, if you narrow down the question well, light research works. And here comes the important point: this approach fits with a logic that large institutions already apply in other areas: decisions based on evidence and input from affected groups, not just internal impressions.
These methods do not compete with a large survey when you need statistical representativeness. They compete with the real alternative, which is usually: “we don’t know, we decide anyway”.
What you are looking for here is sufficient evidence to adjust communication decisions. And, if you can, triangulation: two or three different sources pointing in the same direction.
Short interviews are useful for understanding what people interpret and why. Not to get headlines, but to find patterns of understanding and friction.
A simple format: 6–8 questions, always the same, with room for follow-up questions. If your goal is to improve a page, a message or a process, ask about specific moments:
With 8–12 well-conducted interviews, very clear patterns usually emerge. The important thing is not the perfect number, but the quality of the script and the minimum diversity of profiles (without obsessing over impossible segmentations).
Useful output: a list of 5–8 findings with textual examples (anonymised) and the associated decision (“if X happens, we change Y”).
Many communication problems are not apparent in an interview because the person has already forgotten the details. The friction diary captures the moment: when someone gets lost, gets angry, hesitates, or gives up.
There is no need to ask for a long diary. You can do it in a simple format:
The idea is to gather information about what they were trying to do, what stopped them, and what they did next. This gives you valuable insights for adjusting the order of information, language, navigation, or microcopy.
Useful output: a map of friction points by “moment” (start, middle, end) and by channel (web, in person, telephone).
When we talk about rumours, many people think of “monitoring networks”. And that’s where fear kicks in: privacy, surveillance, bias, political reaction.
There is another way. Rumour analysis, when done properly, need not invade private conversations. It can rely on already available signals:
If you also use social listening, it is advisable to do so with clear ethical criteria. The World Health Organisation (WHO) offers a specific webinar on social listening in managing “infodemics” (rumours and misinformation in emergencies), emphasising an ethical approach, privacy and respect for human rights.
Although the context is healthcare, the logic is applicable: listen to understand and mitigate damage, not to monitor.
Useful output: a “living list” of 10 rumours/misunderstandings, with probable origin (information gap, ambiguity, bad experience), most affected group, reputational risk, recommended response (clarification, process improvement, spokespersons, frequently asked questions).
This method is very effective and minimally invasive if done tactfully. It is not about recording anyone without permission. It is about observing how they interact with content or a channel in a controlled environment:
You look at where they hesitate, where they go back, what word blocks them, what they ignore because it is hidden.
Observation helps you avoid a common pitfall: “as I understand it, it is understood”. In institutional settings, this pitfall can be costly.
Useful outcome: 10 “blocking points” with proposals for immediate adjustment (order, title, clarification, example, call to action).
Microgroups are useful when you need contrast and nuance, especially on sensitive issues. We are not talking about a large, showy “focus group”. We are talking about 4-6 people, 45 minutes, and a well-formulated question.
A microgroup works if:
Here, the role of the facilitator is key: to prevent one voice from dominating and the group from turning into a political debate with no way out.
Useful outcome: drafting and framing decisions based on real language, not on the team’s assumptions.
Here is the leap that differentiates applied research from “decorative research”: the step from finding to decision.
A practical way to do this is always to use this structure:
Finding → Impact → Decision → Minimum evidence → Responsible party → Review date
This also protects you internally: it avoids endless discussions about “opinions” and puts you in a framework of continuous improvement.
If you think about it, it is the same logic that is defended in institutional approaches to better regulation: design and evaluate with evidence, incorporate stakeholder input, and use that material to improve decisions.
1) Clarity decisions (language and structure)
Changing titles, ordering information, adding examples, separating “the essential” from “the complementary,” and avoiding ambiguities. This is usually the cheapest and most cost-effective change.
2) Channel and pace decisions
It is not just a matter of “publishing in more places.” It is about deciding where each thing should happen: what goes on the website, what goes to citizen services, what is answered with a template, and what requires conversation.
3) Internal coordination decisions
Many “communication crises” are coordination failures. If research shows that different areas respond differently, the decision may be to unify criteria, record an official response, and define who validates it.
If you are investigating based on citizen participation or contributions, it is not enough to “collect”. You have to respond. The Organisation for Economic Co-operation and Development (OECD) has pointed out that many governments still fail to close the consultation cycle with public responses on how opinions were incorporated, and that less than half do so consistently.
Responding is not an “extra”. It is part of the process’s trust and legitimacy.
Applied research in institutions has a red line: do no harm to participants, vulnerable groups, or the team itself (which may also be exposed).
Here are some specific practices that work in almost any context.
If you conduct interviews, diaries, or microgroups, consent becomes easy when you are transparent:
The same applies to social listening: define the objective (detect misunderstandings to correct them), the scope (public or aggregated content), and what you will not do (identify individuals).
The WHO webinar on social listening, which I referred to in a previous section of the blog, emphasises precisely the ethical approach and respect for human rights as the basis for these practices.
In European environments, there is also a key distinction: pseudonymisation is not the same as anonymisation. The European Data Protection Board points out that pseudonymised data remains personal data if it can be attributed to a person. That pseudonymisation reduces risks and helps to comply with obligations, but does not automatically make the data anonymous.
Translated into everyday terms: if you store audio recordings, notes or quotes, treat that material as sensitive. And if you don’t need it, don’t keep it.
There is no such thing as 100% “bias-free”. There is “less bias” and being aware of where bias creeps in.
Two practical rules:
In institutional communication, investigating can touch nerves. Not because it is wrong, but because there is political tension, polarisation or fear of “giving ammunition”.
Here, caution is key:
This allows you to investigate without the team feeling like they are opening Pandora’s box.
Applied research is not an academic luxury. It is a tool for better decision-making, with less blind intuition and clearer signals. And, if done carefully, it also reinforces something worth its weight in gold in institutional settings: the perception of fair treatment.
I would like you to explain on LinkedIn what holds you back most from researching at your institution: time, permissions, fear of bias, data, or simply not knowing where to start. I look forward to hearing from you.
Institutions keep designing messages for public feeds while their audiences make decisions in WhatsApp groups no institutional algorithm can reach. The problem is not technical. It is a question about how trust works when public visibility disappears and what remains are relationships.
When a deepfake involving your organisation surfaces, the first instinct is to treat it as a technical emergency. But what determines the real damage is not the fake content itself — it is what your organisation does, or fails to do, in the hours that follow. That is a communication decision.
Is your team already using AI without clear rules in place? This practical guide shows how to build an AI editorial policy for communication: what to allow, who reviews it, and how to disclose its use without putting trust at risk.