Case Study: An Insurance Company Used Clever Psychological Segmentation to Reduce Costs-per-Acquisition by 49%

An Insurance Company Used Clever Psychological Segmentation to Reduce Costs-per-Acquisition

In this case study, you’ll discover:

  • The story of a campaign that took off once it landed on Facebook;
  • When to use object-focused rather than people-focused product descriptions; and
  • What type of people don’t respond well to targeting and why you shouldn’t waste time and money trying to convert them. 

Picture a young female living in the US. Let’s call her Naomi. Naomi is agreeable and slightly neurotic  – a good planner if you will. She happens to own a house, which comes with a lot of things that can cause grief…such as…plumbing.

case studies square banner
Get access to all case studies available on InsideBE

Discover ground-breaking ideas and fascinating solutions.


But Neurotic Naomi knows what to do when something breaks; and that’s to reach out to someone she knows (e.g. a handyman, friends, or family) for help. When choosing whom to trust, she doesn’t mess around – Naomi relies on customizability and comprehensiveness and is influenced to buy based on good reputations. 

It’s no wonder authoritative sources do it for her. Church is important as an influencer, as are websites (reviews, price comparison). Above all, she values a solution that’s reliable and stress-free. 

Authority Bias

Authority bias is the tendency to attribute greater accuracy to the opinion of an authority figure and be more influenced by that opinion.

Source: wikipedia

Now picture a company that has a solution for Naomi. It sells utility insurance to homeowners and services millions of contracts across most US states. As a homeowner, you are protected from unexpected costs and hassle when something goes wrong unexpectedly, with 24/7 support and prompt, professional service.

It sounds exactly like something Naomi might be willing to try as it would give her peace of mind, which she values above all else. But the thing is, she’s never heard of them!

In the past, this company would have sent her a direct mail with a high cost-per-acquisition. What a bummer! And the second bummer? Naomi would not have even been on the mailing list! That’s why the company was in desperate need of a new strategy.

3 Principles of Behavioral Science That Make Customers Act
  • Article
Case Study: Fundamental Mistakes in Email Communication That Impact Asana’s Upselling Efforts

And thanks to Patrick Fagan (behavioral scientist and now Chief Scientific Officer at Capuchin Behavioural Science) and the team he worked with at the time, the company figured out that what would resonate with Neurotic Naomi (and perhaps others like her across the United States) the most is both reassuring and soothing messaging that would calm her fears and show how to prevent negative occurrences. 

The insurer also understood how crucial it was to present themselves as a reliable friend that Naomi knows and can rely on while still maintaining their authority and persuading via authority appeals that she would appreciate. They also knew that she’d be more likely to opt for utilities insurance after she had heard about all the cost benefits of the warranties. 

The insurer also understood how crucial it was to present themselves as a reliable friend while still maintaining their authority.

So they would hit her with a Facebook post, which would assure her that she could avoid a nasty pipe burst and the expensive repairs and actually protect her pipes with an affordable in-home plumbing emergency protection program.

It would also be crucial for Naomi to learn that this service allowed for a monthly payment which wouldn’t cost her an arm and a leg and also that a company had a 24/7 hotline she could call even at 2 am. This post, which Naomi and other users like her saw, has saved the company $27.04 on cost-per-acquisition.

masterclass course thumbnail
  • Online course
Behavioral Economics & Psychology in Marketing (with certificate)
learn more

But the story doesn’t end there. Apart from Naomi, there were 5 other segments hit during a revolutionary, targeted digital advertising campaign, which ran on Facebook over the course of a month and reduced cost-per-acquisition by up to $29.35 (a massive 49% reduction).

But let’s take it one step at a time. At the end of it, you’ll have learned what type of customer might be a dead-end, where to focus your efforts instead, and specific copy for you to hit each segment. 

The implicit research 

The first step Fagan’s team took was to run a psychometric segmentation survey on over 2,000 American homeowners. It was sent to their customer list but also to an online panel of prospective customers (those who could conceivably buy this type of insurance). 

The survey mapped out personality traits, interests, motivations, and barriers to buy (the latter two of which are actually a key questions for segmentation) as well as more intricate details about messaging preferences such as tone of voice and media content preferences.

Since Fagan specializes in uncovering true consumer responses and behaviors rather than what consumers claim, this subject is close to heart. The key point he makes is that instead of asking about attitudes, it focused on behaviors and also used projective indirect questioning techniques.

Fagan points out that instead of asking directly “What do you think about utilities insurance?”, you can ask in a more indirect way: “If utilities insurance were a person, who would it be?” If they say their mother, then there’s some insight you can get from that, according to Fagan. 

“If utilities insurance were a person, who would it be?” If the customer says it would be their mother, then there’s some insight you can get from that, according to Fagan.

Another way to get to the bottom of what people actually think (rather than what they declare), is to ask what images they associate with a given product/category. If someone associates insurance with a leaking pipe (of money), it gives you an idea of how useful (or not) it is for them. 

Fagan, who uses established psychometric models like The Big Five, also recommends doing qualitative research and literature review before segmentation: “You can search Google Scholar for things like motivation for buying insurance and find a lot of useful academic research,” he explains. 

You can also run a mini survey to find out what nudges people respond best to. For example: If there were 6 insurance options available — all else being equal — which one would you buy? The most popular, which uses the principle of social proof? The one that gives me a free gift (a concept known as reciprocity)? And so on and so forth… 

In this case, the results of the survey informed the subsequent strategic and creative recommendations for each segment (i.e., nudges, aesthetics, proposition), which were then executed and targeted to the right segments on Facebook.

Optimizing Nudge Impact with Consumer Neuroscience Insights
  • Article
Case Study: How Ogilvy Used Neuroscience to Test the Name and Packaging Design for NanoSalad

Bespoke posts for different segments

Apart from Neurotic Naomi, whom you’ve already met, there were 5 other segments that differed in motivations and barriers, each of which had their own post. Let’s take a look at them now.

  1. Shrewd Shawns have experience with the brand and positive perceptions of it (easy to buy, low cost, etc.). Shawn has experience with warranties and sees them as a means of protection and assistance. He values a more professional approach; the most important influencer is the provider or nobody. When something breaks, he contacts the insurer/warrantor. 

What will resonate well with a Shrewd Shawn is: 

  •  Calming and reassuring messages that focus on the protection and help that warranties provide.
  •  References to the company’s solid reputation and its excellent customer service. 
  • Focusing on what could go wrong if one doesn’t have a home warranty, and how the company could easily prevent that (loss aversion).
  • Using persuasive appeals around urgency, scarcity, and authority.
  • Using muted, cool colors that are preferred by neurotic individuals.

Fagan’s team suggested targeting them with reassuring messages, which reference the company’s excellent reputation combined with family, community oriented visuals. 

The Facebook ad post should point to the company’s longevity by stressing out that it has been providing homeowners with high-quality service for X years.  Showing commitment to helping families take care of their homes as well as mentioning high customer approval rate would also be key for this segment.  

  1. Handy Hanks like to do things themselves. They aren’t really into warranties and don’t know much about them. If anything, a Handy Hank thinks that while warranties might help mitigate the inconvenience of having to fix things, the repairs aren’t always as good.  He doesn’t think much of authority figures and wants info straight from the source. A handyman is an important influence.

When making a purchase, a Handy Hank wants the cheapest option and doesn’t care about the other options.

What will resonate well with a Handy Hank is: 

  • Creative, novel, and even unusual messaging and aesthetics.
  • Encouraging them to think for themselves (e.g., using metaphors and rhetorical questions).
  • Illustrating how warranties can help them help themselves and be self-sufficient.
  • Focusing on the straight-up facts and being utilitarian.

For this segment safeguarding their home is key. So they hit Handy Hank with a straightforward message to encourage reflection and show him how to be. The post and copy vividly described the outcome they didn’t want; the visuals showed actual things that could go wrong in the house such as broken sewer line, water pipes burst, and the outcome such as a faulty outlet or a panel. The copy also stressed out their no-nonsense approach such as comprehensive protection, services, and support.

masterclass course thumbnail
  • Online course
Behavioral Economics & Psychology in Marketing (with certificate)
learn more

3. Sociable Sandies are all about family and close relationships. She’s familiar with the company or has used it before. She thinks it’s old-fashioned and inexperienced. However, that being said, she thinks warranties provide a good peace of mind.

If something happens, Sandie won’t fix it herself but will turn to family or friends for help. So it’s no wonder her most important influencers are parents, siblings, cousins, etc. Purchases made via phone or in-person are important to her as Sandie wants sociable relationships with trusted people.

What will resonate well with a Sociable Sandie is: 

  • References to the company’s solid reputation for quality repairs.
  • Communicating that it’s a company trusted by friends and family members.
  • Showing that warranties provide peace of mind.
  • Creative, novel, and even unusual messaging and aesthetics.
  • Being aspirational and focusing on how warranties can make one’s life better.

Fagan’s team targeted her with creative and unusual messaging that painted the company as both a trustworthy and high-quality insurance provider. 

The Pinterest-like abstract visual with no tagline was accompanied by the post which encouraged Sandie to not stress over unexpected costs in case she had a service line problem. The key message was to convey she would get peace of mind and help any time and that the whole process would be hassle-free.

Social Proof: The Undiscovered Power and Hidden Pitfalls of This Well-known Principle
Download now

4. Reserved Reginas don’t have a good view of warranties and are reluctant about the service. They think it helps you fix things and protects you from unexpected bills, but it’s not worth the money. The most important influencer for a Reserved Regina is a realtor or mortgage broker and she is less likely to listen to handymen or friends and family.

She can be influenced by familiarity with an established provider, speed, and the ease of making a purchase, plus comprehensive coverage.

What will resonate well with a Reserved Regina is: 

  • Emphasizing the cost-effectiveness of warranty ownership.
  • Being aspirational and highlighting the potential rewards to be gained rather than the pitfalls to be avoided.
  • Using warm, vibrant, and diverse aesthetics that incorporate social scenes.
  • Being friendly but authoritative and formal in language and tone.

5. Hermit Herberts aren’t influenced by anything or anybody, especially not by their church. They work out what the problem is and do it themselves. A Hermit Herbert will have no opinion of the company because he’s never heard of them.

What will resonate well with a Hermit Herbert is: 

  • Being factual and utilitarian.
  • Looking to ‘rebellious’ genres for inspiration, such as horror films, comic books, or punk rock.

The outcome: a saving of 29.35$ per acquisition

Once these profiles were created, the company uploaded email addresses from their existing customer base on Facebook and did a lookalike model where they found similar people based on the things those email addresses liked and then targeted the ads to them. 

The digital campaign resulted in a cost-per-acquisition that was up to $29.35 lower than their direct e-mail approach. However, two segments performed worse than their baseline and they were Hermit Herbert and Reserved Regina.

“Hermit Herbert was very much a naysayer segment, probably more an artifact of people who say “no” to every question, so we may not have been targeting a genuine segment. Alternatively, they might just be hard to convert in general if they’re so negative and tend to disagree” Fagan explains. 

As for the Reserved Regina, it was a segment that didn’t know much about the brand/category, so they were just quite hard to convert in general. 

Fagan points out that in both cases, it would probably be better not to target these people at all as they’re low priority segments: “Normally in segmentations, I would say which segments are a priority or not, butin this case the client wanted to try all of them,” he adds.

Key Takeaways:

  • In order to tease out how people view your product, instead of asking them directly “What do you think about the product?”, ask them in a more indirect way: “If this product were a person, who would it be?” Alternatively, you can ask what images customers associate with the product.
  • If you aim to create bespoke posts and visuals for each of your segments, be smart about it. Don’t pour your effort into a naysayer segment as they’re just plain hard to convert in general. Focus more on the low hanging fruits like past users or those with brand awareness.
  • Run a mini survey to find out what nudges people would respond best to.
  • You don’t have to go “all or nothing.” Even if you can’t create a separate message for each segment, combining them into a single message is not a bad idea. Read more on how to do that here.

Introduction of an expert
  • Case study by
  • Patrick Fagan
  • Behavioral Scientist, Consultant