Web app · Communications and PR

MessageLab

Know how a message lands before it goes out.

The MessageLab feedback screen on an angled MacBook Pro: a slogan about vaccines and four AI audiences reacting to it, from Love it! to Poor
All projects

MessageLab lets PR, political and communications teams test a slogan, statement or document on AI personas and cohorts. Each audience answers with a reaction, a quote and the reasons behind it. We designed the product's UX/UI.

Our role
Product UX/UI design
Industry
Communications and PR
Platform
Desktop web app
Year
2025

01The challenge

A message gets one chance in public.

Comms teams write for people they rarely get to ask: voters, policy makers, clients, a whole country. Polls and focus groups take weeks. The product had to give them a fast, honest read before launch.

  • Feedback before launch.

    Teams need to know how a line will land with each audience, and what makes it fall flat, while there is still time to change it.

  • The right audience.

    A message for Dutch teachers is not a message for EU influencers. Finding the right personas and cohorts in a large library has to be quick and precise.

  • Feedback you can act on.

    AI answers are only useful if a strategist can read them at a glance, ask why, and share the result with the team or the client.

02Our approach

Built around one test run.

We designed the product around the loop every comms team repeats: write the message, choose who hears it, read what they think, improve it.

  1. One guided workspace.

    The left panel holds the whole test in order: message type and text, the audience, an optional framework such as Cialdini's principles, then Test Message.

  2. Feedback in layers.

    Results use progressive disclosure. The first view shows one reaction and one line per audience. Pick a persona to open their reasons and highlights, and chat with them only if you need to know more.

  3. Personas as real profiles.

    Each persona is a person with a face, a short bio and attributes grouped the way strategists think: demographic, culture, socio-economic, professional, psycho-social and political views.

03The product in action

A week at Northbridge Comms.

GET STARTED

From sign-up to the first project.

Sign up with email or Google, confirm the code, then tell MessageLab how big your team is and what you want to test. A short form creates the first project and the workspace is ready.

TEST A MESSAGE

Write it, pick the audience, generate.

Type a slogan, statement or document, choose personas and cohorts from the library, and add a framework if you want one. One click later, each audience has answered.

FROM OVERVIEW TO DETAIL

See the big picture, then go deeper.

AI feedback can be long, so results open in layers. After Test Message, each audience shows one reaction and one line. Pick a persona to see their full answer and quotable highlights, then chat with them to ask why. When the story is clear, share a link with the team or the client.

PERSONA LIBRARY

Personas you can trust.

Every persona has a face, a short bio and attributes grouped by theme. Edit them step by step, or add your own to the library and keep favourites in folders.

Two MessageLab persona screens: Emma Johnson's profile with her demographic attributes, and the edit persona form with her picture, name and brief

04Design language

A dark frame, a bright page.

A deep navy frame holds the tools, and the work sits on a white page in the middle. One coral accent marks the next step. DM Sans keeps long feedback easy to read, and real faces make every audience feel like people.

The MessageLab style tile: navy and coral colours, DM Sans type samples, a persona feedback card, the Love it, Not bad and Poor reactions, a persona card from the picker, the message type switch and the audience demographics

05The result

A second opinion from every audience.

We designed the MessageLab web app: sign-up and workspace setup, projects, the message workspace with persona feedback, chat and sharing, the persona library and the account and team settings.

Next projectBlueBubbleStudent money, one clear answer a day.

06Contact

Building an AI product?

We design apps and platforms from first idea to next release.

  • No commitment, just a conversation
  • You talk directly with a designer
  • A reply with a clear next step
Prefer to talk? Book a call

Send us a message

Tell us a little about your product. We’ll get back to you soon.