Product Messaging

Most financial product messaging is written inside-out — starting from the product’s features and working outwards. Dialogue lets you work the other way: start with how real people describe their problems, fears, and goals, then build messaging that meets them there.

This guide walks through building a positioning brief for a SIPP product targeting mid-life accumulators.

Step 1: Map the emotional landscape of your target segment

Start by understanding what your target segment is actually saying — not what you assume they care about.

$
GET /api/v1/comments?life_stage=mid_life_accumulator&product=pension_sipp&min_likes=5&limit=50

Read through these comments before doing anything else. You’re looking for:

• The words they use to describe the problem
• What they’re afraid of
• What they wish they’d done differently
• What would make them feel in control

This qualitative read is the foundation of any credible messaging framework.

Step 2: Identify what triggers engagement with the category

Find the moments that make people pay attention to their pension for the first time — or re-engage after years of ignoring it.

$
POST /api/v1/search
$
{
>
  "query": "what made me finally start thinking seriously about my pension in my 40s",
>
  "limit": 15,
>
  "threshold": 0.4,
>
  "life_stage": "mid_life_accumulator"
>
}

$
POST /api/v1/search
$
{
>
  "query": "realised I hadn't looked at my pension in years",
>
  "limit": 15,
>
  "threshold": 0.4
>
}

Common triggers you’ll find: a round birthday, a colleague retiring, a health scare, redundancy, or simply stumbling onto a personal finance video. These trigger moments are where your media targeting and opening messaging should live.

Step 3: Find the language of the core problem

The gap between what people have saved and what they need is the central tension of pension marketing. Find out exactly how your segment describes it.

$
POST /api/v1/search
$
{
>
  "query": "not saved enough for retirement feel behind",
>
  "limit": 15,
>
  "threshold": 0.4,
>
  "life_stage": "mid_life_accumulator"
>
}

$
POST /api/v1/search
$
{
>
  "query": "don't understand my pension confused about what to do",
>
  "limit": 15,
>
  "threshold": 0.4
>
}

Pay close attention to the specific vocabulary. Do they say “behind”? “Scared”? “Confused”? “It’s too late”? Those words belong in your messaging — not “maximise your retirement income potential.”

Step 4: Pull the most resonant proof points

High-engagement quotes from people who have taken action — and felt better for it — are your most powerful messaging material.

$
GET /api/v1/quotes?life_stage=mid_life_accumulator&product=pension_sipp&emotional_register=peace_of_mind_relief&min_likes=10&limit=10

These are the “after” stories. People describing the feeling of finally understanding their pension, sorting their contributions, or getting a forecast they could trust. They’re the emotional destination your messaging should point towards.

Step 5: Generate a positioning brief

With the raw material gathered, use the AI analyst to synthesise it into a structured positioning framework.

$
POST /api/chat
$
{
>
  "messages": [
>
    {
>
      "role": "user",
>
      "content": "Build a product messaging framework for a SIPP targeting mid-life accumulators aged 40-55 who know they're behind on pension saving. Include: the core tension, the key insight, three positioning territories each with a strategic rationale and example message, the language to use and avoid, and the emotional journey from problem to resolution. Ground everything in real consumer language from the corpus."
>
    }
>
  ]
>
}

Using the MCP server

In Claude Desktop, the full research and synthesis happens in a single conversation:

“I’m positioning a SIPP product for people in their 40s who’ve been neglecting their pension. Pull the data from the corpus — what does this segment fear, what triggers them to act, and what language do they use to describe the problem? Then give me three positioning territories with example headlines.”

Follow up with:

“Which of those territories has the strongest evidence in the corpus? Show me the quotes.”

Building a messaging matrix

Once you have your territories, use the API to test each one against the corpus:

$
POST /api/v1/search
$
{
>
  "query": "[your proposed headline or value proposition]",
>
  "limit": 10,
>
  "threshold": 0.35,
>
  "life_stage": "mid_life_accumulator"
>
}

If you get strong semantic matches with high-engagement comments, the territory has evidence. If results are sparse or low-relevance, the message may be based on assumptions rather than real consumer concerns.

Tips

• Compare emotional registers across segments — the same product (pension) hits completely differently for young_pre_property (abstract and distant) versus pre_retirement (urgent and frightening). Use life_stage filtering to build segment-specific messages.
• Validate with signals — comments tagged implicit_deliberator are people actively considering a product decision. These are your highest-intent voices and the most useful for conversion-oriented messaging.
• Test your jargon — search for terms you’re considering using (e.g. “drawdown”, “annuity”, “lifestyling”). If you find them used confidently, they’re in-vocabulary. If the surrounding comments express confusion, they’re jargon.