Case study · Health mutual insurance

How Harmonie Mutuelle tripled its email open rate

A behavioural approach applied to segmentation and campaign design, deployed through A/B testing waves, on both corporate and retail segments.

Context

Harmonie Mutuelle is one of France's leading health mutual insurers. It operates in highly competitive markets, where acquisition costs are under constant pressure. Its email campaigns had historically relied on the sector's classic playbook: detailed coverage descriptions, tariff comparisons, technical language.

On difficult segments (families, corporate market), open rates plateaued. Classic UX optimisations (send time, subject length, first-name personalisation) had stopped moving the needle.

The problem

The issue was not the channel, but the grammar of the messages. Research on information processing under uncertainty (Kahneman 2011, Sunstein 2014) shows that a recipient sorts their inbox in a few hundred milliseconds using System 1, not through analytical reading. Leading with coverage details demands a cognitive effort that most recipients will not engage.

Above all, a recipient who already believes they are healthy systematically underestimates their future risk (unrealistic optimism, Weinstein 1980). A campaign that hammers on worry hits a wall. A campaign that activates the right behavioural levers goes around it.

Our approach

We applied two complementary frameworks.

MINDSPACE (Dolan, Hallsworth, Halpern, King, Vlaev 2010), to rewrite each message by activating the nine behavioural levers validated by the Behavioural Insights Team: Messenger, Incentives, Norms, Defaults, Salience, Priming, Affect, Commitments, Ego.

COM-B (Michie, van Stralen & West 2011), to diagnose what the recipient lacks to open and act: capability (does she know what to do?), opportunity (are channel and timing right?), motivation (have we connected the message to a benefit that matters to her?).

Concretely, the work unfolded in three phases.

Behavioural framing by segment. For each segment (families, corporate market, retail with health difficulties), we identified the dominant biases and specific psychological frictions. A SME executive does not decide like the parent of a large family, and a self-employed worker decides like neither of them.

Co-design workshops with Harmonie Mutuelle's marketing teams. Rewriting subject lines, hooks and arguments while explicitly activating the right MINDSPACE levers for the right segment. The teams worked with an operational grid, not with an awareness seminar.

A/B testing waves, with continuous iteration. Each variant tested against a control version, learnings capitalised, best version deployed, then a new wave.

Results

On the corporate segment, the open rate was multiplied by three, moving from around 10% to a plateau of 23 to 36% depending on the wave.

On retail segments, open rates stabilised between 60 and 70%, whereas insurance/mutual health benchmarks typically sit between 18 and 25% (DMA Insight Email Benchmark 2024).

Beyond the numbers, the methodological transfer: the behavioural grid was integrated by the marketing teams into their standard design cycle.

What this case illustrates

Behavioural levers do not replace a marketing strategy: they amplify it. On a mature channel like email, gains do not come from technical optimisation but from rewriting in behavioural grammar. Tools are available: MINDSPACE and COM-B have each been around for more than fifteen years. What most teams lack is disciplined, segment-by-segment application.

Frameworks used

MINDSPACE, COM-B, unrealistic optimism (Weinstein 1980), social proof (Cialdini), System 1 / System 2 (Kahneman 2011). See the glossary for definitions.

Frequently asked questions

Which behavioural frameworks did Krakn use for Harmonie Mutuelle?

MINDSPACE (Dolan et al. 2010) to structure message rewrites, and COM-B (Michie et al. 2011) to diagnose what the recipient lacks to act. Two complementary frameworks: MINDSPACE works on the message angle, COM-B works on the conditions for action.

How do you measure the impact of a behavioural marketing campaign?

Through systematic A/B testing. Each behavioural variant is tested against a control version, on samples representative of the segment. Primary KPIs are open, click and conversion rates. Secondary KPIs cover lead quality and acquisition cost.

On which segments did Harmonie Mutuelle achieve the strongest results?

The strongest relative gain was on the corporate segment (open rate multiplied by three). The highest absolute level was reached on retail segments (60 to 70% open rate).

How does MINDSPACE differ from COM-B?

MINDSPACE is a checklist of nine behavioural influence levers (Messenger, Incentives, Norms, Defaults, Salience, Priming, Affect, Commitments, Ego), useful when designing messages. COM-B is a behavioural diagnostic model that breaks any action into three components (Capability, Opportunity, Motivation). MINDSPACE answers "how to phrase", COM-B answers "why the recipient does not act".

Can this method be applied to sectors other than health insurance?

Yes. MINDSPACE and COM-B have been validated across domains as varied as public health, taxation (HMRC UK), road safety, personal finance and organ donation. Krakn's method applies these frameworks to any sector where consumer decisions are hindered by documented cognitive biases: typically insurance, banking, energy, mobility and health. Want to explore the same work on your campaigns? Let's talk →

Ready to decode your clients?

Let's talk about it.

30 minutes to see how behavioural science can transform your business.