How one insight reshaped a market entry → and what any company can steal from the process

The brief that shouldn’t have been possible

Lola&Lykke is a Finnish breastfeeding company → genuinely well-engineered pumps. Five components to clean instead of twelve. Quiet enough to use at 3am. Simple enough for anyone in the household to set up without reading a manual. The products had traction in the UK and other markets. Sweden should have been a natural next step.

It wasn’t simple. No Swedish customer base. No local team. No significant budget. And a target audience → parents expecting babies in April 2026 → who were going to give birth before a traditional research agency could deliver page one of a findings deck.

Laura McGrath, CEO, faced a decision that confronts every startup entering a new market. The traditional route → hire a research agency (€15,000-€40,000, eight to twelve weeks), a brand consultancy for positioning, a comms specialist for voice, a community strategist for activation, and a project manager to stop them contradicting each other → was impossible. The total: €40,000 to €100,000, three to six months. For a startup with constrained resources and a window measured in weeks, that’s fiction.

The alternative → skip the research, launch on instinct, find out what the audience thinks six months later after the budget is gone → was equally unacceptable. That’s how you spend your marketing budget learning lessons you could have learned in advance.

Laura chose a third option. She brought in a single strategic partner → Aizle → to do all of it. One person. Six weeks. Research, positioning, voice development, market activation, and community architecture.

The investigation

I built four detailed behavioural profiles representing distinct Swedish parent archetypes. Not demographic cutouts → psychological portraits.

Johan, 34, a software developer in Stockholm on paternity leave. Technically competent, emotionally invested, frustrated by equipment complexity that prevented him from contributing to feeding. His internal monologue included thoughts he’d never say aloud: “I feel useless when she’s feeding. I just sit there.” He’d research pumps obsessively → building mental comparison matrices, reading spec sheets, cross-referencing reviews on Prisjakt → because the research process itself made him feel like he was contributing. He wanted equality in parenting but felt guilty about wanting recognition for it. That contradiction meant he’d respond positively to “team sport” messaging on the surface → but the positive response masked a more complicated emotional reaction that only showed up under stress-testing.

Elin, 36, a project manager on her second child. She’d hated her first pump → too many parts, too loud, felt like medical equipment rather than something designed for a human being. She knew exactly what she wanted this time. She actively distrusted influencer content and would post honest reviews on parenting forums. She projected total confidence. Underneath that, she was running on four hours of sleep and silently furious at how much of the parenting load still fell on her. That tension → the gap between the competence she projected and the exhaustion she felt → meant that messaging which praised her coping skills landed as tone-deaf. She didn’t want to be told she was doing great. She wanted a pump that didn’t add to her workload.

Klara, 28, a first-time teacher. Anxious, authority-dependent. She would buy whatever her BVC nurse recommended because she lacked the bandwidth to research independently. She privately suspected her nurse just recommended the safe default rather than the best option → but she couldn’t muster the energy to challenge it. Her trust hierarchy was the steepest of all four profiles → healthcare professionals at the top, peers and forums in the middle, brand messaging at the bottom. An Instagram ad wouldn’t reach her. A recommendation from her midwife would.

Sara. The most complex profile. A parent who is silently struggling → where the feeding problem masks something deeper. Where exhaustion has crossed into something that might need professional support, but she’s too tired to seek it. Where the wrong message from a brand → a cheerful “you’ve got this!” → isn’t encouragement. It’s gaslighting. Sara was the profile that stress-tested every piece of messaging against the worst-case scenario: if someone having the worst day of their life reads this, does it help or does it harm? Most category messaging fails the Sara test. Badly.

Each profile included internal contradictions, because real people are inconsistent. Johan wants equality but craves recognition. Elin projects confidence while running on empty. Klara defers to authority while privately doubting it. Sara needs help but can’t ask for it. Those tensions → the gaps between what people show and what they feel → are where the strategic insight lives. A demographic segment can’t tell you about those gaps. A behavioural profile can.

I stress-tested Lola&Lykke’s positioning across five structured exercises → tone calibration, competitive positioning, cultural credibility, upgrade framing, and an adversarial round that deliberately attacked the strongest positioning to find where it breaks under pressure. The whole process took days, not months.

The findings arrived fast. Several directly contradicted what the client assumed about their audience.

Good. Research that confirms what you already believe is expensive reassurance. Research that surprises you is worth paying for.

The reveals

The tone.

Swedish parents unanimously rejected both ends of the spectrum. “We understand that breastfeeding can sometimes feel challenging” → corporate hedging, forgettable before the reader finishes the sentence. “Pumping sucks, let’s stop pretending otherwise” → reckless in a category where some readers are genuinely struggling and a flippant tone lands as dismissive of their pain.

The sweet spot was a narrow, specific register we called Level 2. Honest, practical, warm without being patronising. “Pumping is hard. We make it easier. That’s it.” Not a slogan. A calibration. Miss it in either direction and you lose someone. Level 2 was the only tone that worked across every single behavioural profile. The only one.

The toxic phrase.

“You’ve got this, mama!” was not mildly ineffective. It was the single biggest negative trigger in the entire study. A categorical rejection of the industry’s dominant emotional register. Lola&Lykke was about to build their entire Swedish communications around it. That one finding, surfaced in the first week, was worth more than the entire engagement cost.

The trust hierarchy.

For first-time and vulnerable Swedish parents, healthcare professionals → BVC nurses, midwives → are the primary influence on purchasing decisions. More trusted than peer reviews. More trusted than advertising. More trusted than anything a brand can say. This finding reshaped the entire go-to-market approach. The client was planning brand-to-consumer communication. They needed to talk to the people consumers already trust.

The fault line.

Positioning breastfeeding as a “team sport” → shared responsibility between parents → was one of the most polarising messages in the study. Johan responded enthusiastically. Elin found it invasive → a brand prescribing how her relationship should work. This wasn’t a problem to solve. It was a strategic decision to make. The research recommended leading with product capability (“built for anyone to use”) rather than relationship framing, letting parents assign their own meaning.

The inclusive upgrade.

“The pump you wish you’d had the first time” resonated with second-time mothers but excluded first-timers. “For when you know what you really need” captured both segments without exclusivity risk → and preserved word-of-mouth potential, because experienced mothers would recommend it to first-time friends without creating a hierarchy.

“Several findings directly contradicted our assumptions, and every piece of strategy we’ve built since has been grounded in that research.”

Laura McGrath, CEO Lola&Lykke

What this method can and can’t do

I want to be honest about this, because overclaiming is how methodologies lose credibility.

Behavioural profiles cannot surprise you the way real humans can. They reflect patterns in training data → they can tell you what plausible people might think, not what specific real people will think. A positive reaction from a behavioural profile is weak evidence → it means the positioning is broadly compatible with how this archetype is likely to respond. A negative reaction is much stronger evidence → if a well-constructed profile recoils at your messaging, a real person almost certainly will too.

The method is better at invalidation than validation. It is excellent at killing bad ideas cheaply. “You’ve got this, mama!” didn’t need a €30,000 focus group to be identified as toxic → but without the behavioural stress-testing, it would never have been questioned at all. The method sits in a space that didn’t previously exist for most startups: between “guess and hope” and “spend €30,000 and wait three months.” It makes rigorous strategic thinking accessible to companies that assumed they couldn’t afford it.

The findings are hypotheses worth acting on, not certainties. The Level 2 tone calibration was validated unanimously across all four profiles → that’s a high-confidence signal. The Team Sport fault line was a genuine split → that’s a strategic decision to make, not an answer. The research tells you where the confidence is high, where it’s low, and where you need real-world validation. The Mum Collective → the community we built → is that real-world validation. The virtual research generates the hypotheses. Real humans confirm or challenge them.

This distinction matters because it’s the honest version of what AI-powered research can do. And honestly naming the limitations builds more trust than pretending they don’t exist.

What was built → and why each decision traces back to a finding

Every deliverable was a direct consequence of the research. Not a separate workstream. A consequence.

The research found that tone calibration was the highest-impact decision → so the first downstream deliverable was a Social Voice Guide. Not a branding exercise. A risk management tool disguised as a style guide. Word choice tables that replaced vague principles with specific decisions: instead of “our revolutionary breast pump,” write “a pump with five parts to clean, quiet enough to use without waking anyone.” Eighteen banned phrases, each banned for a specific reason the research surfaced. Structural guidelines for social posts, DM templates, and visual direction → because the research was unambiguous that aspirational imagery (white linens, minimalist nurseries, serene smiling) destroyed the credibility of honest copy. The team learned to hear their own instinctive language choices and question them against the research. That skill stayed after the guide was delivered.

The research found that honest, specific language generates honest, specific engagement → so the activation copy for a Swedish parenting group was written using the findings as a blueprint. The result: 78 substantive comments from a single post. Not emoji reactions. Paragraphs. Parents sharing real experiences about breastfeeding, uncertainty, and the gap between expectation and reality. More genuine engagement from one post than the client had generated in their entire previous period of Swedish marketing activity. The team saw, in real time, that the research-derived tone worked → and they understood WHY it worked, which meant they could replicate it.

“All this came from one simple recruitment copy in a Swedish FB parenting group that generated 78 substantive comments from a single post. Not emoji reactions. Paragraphs. Parents sharing real experiences.”

Laura McGrath, CEO Lola&Lykke

Then came the community. And this is where the counterintuitive decisions start stacking up → each one grounded in a specific finding, each one the opposite of what most marketers would do.

Most brands put a marketing hire in charge of their community.

We put an independent lactation consultant. Maria from Mindful Milk → a Swedish-speaking healthcare professional whose expertise made the group genuinely useful to participants. They’d normally pay for her access. Her independence was the costly signal that this wasn’t a sales channel. The research was unambiguous: healthcare professionals are the most trusted voice for this audience. A brand account moderating a group of new mothers is a fox guarding a henhouse → everyone senses it, nobody says it. Maria’s presence changed the psychological contract entirely. She was also trained to recognise when a participant might be struggling and respond with empathy and a referral to professional support → not with product talk. In a community of new mothers, some of whom may be experiencing postpartum anxiety or depression, that’s not a nice-to-have. It’s a duty of care that a marketing hire isn’t equipped to fulfil. The team learned to see moderation as a trust architecture decision, not a staffing decision.

Most brands require content creation, reviews, or social posts as a condition of participation.

We required nothing. Zero content obligations. Participants were explicitly told they didn’t have to post, review, photograph, or promote anything. This is the most counterintuitive decision for most marketers → and the most powerful. When you remove the obligation to create content, you eliminate the transactional dynamic that kills authenticity. The behavioural logic is simple: people are more honest when they don’t feel obligated to be positive. By never asking for reviews, you create the conditions for genuine feedback → including the negative feedback that is more valuable than any testimonial. “The suction on setting 2 feels too strong” is worth more to a product team than a hundred polished five-star reviews. And the paradox → participants who genuinely love the product talk about it anyway, to friends, to their BVC nurse, in the group → because they want to, not because they were asked.

Most brands launch communities with no end date.

We designed a planned lifecycle → a clear beginning, middle, and end, aligned with the natural breastfeeding timeline. Because communities built to run indefinitely don’t run indefinitely. They decay. Posting frequency drops. The community manager moves on. New members trickle in but find a ghost town. Nobody formally closes it because that feels like failure. So it lingers, slowly damaging the brand’s credibility with anyone who stumbles across it. A planned close-out prevents that decay by making the ending a feature, not a bug → the brand commits to sharing what was learned, what changed as a result of feedback, and what ongoing value members receive. That moment of closure is itself a powerful trust signal: the feedback mattered.

Most brands collect community feedback for marketing purposes → testimonials, social proof, content.

We connected it directly to product development. The feedback prompts were designed as behavioural conversation-starters: “You’ve had the pump for a week → what’s the first thing you noticed, good or bad?” and “Has anyone mentioned the pump to their BVC nurse? What was the reaction?” That last question was specifically designed around the trust hierarchy finding → surfacing how healthcare professionals respond to the product, which is insight no other channel could provide. The community feedback now connects directly into Lola&Lykke’s product development pipeline, including two new product lines in active development → a wearable pump targeted for late Q2/early Q3 2026. When feedback shapes real product decisions, participants can eventually see the impact of what they shared. “You told us the suction was too strong, and we adjusted it.” That’s the most powerful thing a brand can say to a community member. It proves the feedback wasn’t performative. The team learned to design for honest input rather than polished endorsement → and to connect that input to product decisions in real time.

The math that should make you uncomfortable

If you’d commissioned this work through traditional channels → research agency, brand consultancy, comms specialist, community strategist, project manager → the conservative estimate is €40,000 to €100,000. Three to six months. And by the time it arrived, the babies would have been born and the window would have closed.

But here’s the part that matters most. A traditional agency would have approved “You’ve got this, mama!” without blinking. Five firms in a room, and not one of them would have had the behavioural framework to stress-test the category’s default language. The expensive option isn’t just slower and more expensive. It’s structurally incapable of finding the insight that mattered most. Not because the people are worse → because the process is designed to confirm existing direction, not challenge foundational assumptions. Nobody gets fired for approving the phrase everyone else uses.

One person. Six weeks. From research to live community.

“We got this for free as Adam’s first case study but are now continuing working with him on a retainer basis, as Adam’s input is simply so valuable to us.”

Laura McGrath, CEO Lola&Lykke

And there’s an irony here that I want you to sit with. This case study is itself a reframing exercise. The same facts told in a different order → discovery first, not chronology first → and the perceived value changes. Same material. Different structure. Different impact. Which is exactly what we did for Lola&Lykke. Same product. Different frame. Different result.

Steal this

These are specific enough to act on, regardless of your industry or product category.

Stress-test your category’s default language before building on it.

Every industry has its “You’ve got this, mama!” → the phrase everyone uses on autopilot, the emotional register nobody questions. Find yours. Test it against your most vulnerable audience segment. The phrase that feels safest might be the phrase that’s doing the most damage. And the costless intervention → just stop saying it → might be the highest-value strategic move available to you.

Map your audience’s real trust hierarchy.

Who do your customers actually listen to when making decisions? Not who you wish they listened to → who they actually trust. Build your activation around THAT person, not your brand account. If the answer is “independent professionals” rather than “influencers,” that changes everything about how you design community, moderation, and outreach.

If you’re building a community: cap it small, moderate with an expert, require nothing, plan the ending.

Twenty-five people who feel heard will generate more honest feedback and more genuine advocacy than five hundred people in a broadcast channel. An expert moderator whose credibility is independent of your brand is the costly signal that this isn’t a sales exercise. Zero content obligations remove the transactional expectation that kills authenticity. And a planned lifecycle prevents the slow decay that turns communities into ghost towns.

Turn your research into tools people actually use while working.

A strategy deck that nobody reads is expensive fiction. A one-page guide with word choice tables and banned phrases that someone uses while writing a social post → that’s infrastructure. The test is: can someone on the team who wasn’t in the strategy meeting use this document to make a better decision? If not, it’s not a tool. It’s a souvenir.

Design feedback prompts as conversations, not surveys.

“What’s the first thing you noticed?” generates a different kind of response than “Rate your satisfaction on a scale of 1-10.” “How would you explain this to your partner in one sentence?” reveals how the product occupies mental space. Mirror how people actually talk about products in their kitchen, not how researchers wish they’d report.

Why Aizle works this way

I launched Aizle to be the thing I kept wishing existed → not an agency, not a management consultancy, but a senior strategic partner who actually does the work. Research, strategy, creative, activation. AI baked into the methodology as the engine that makes rigorous research possible at startup speed. Behavioural science as the backbone of every decision → not intuition dressed up in post-rationalised logic, but actual findings about how people think, decide, and respond.

And education built into every engagement → because the deliverables are the short-term value and the capability is the compound interest. The goal is not that you need Aizle forever. The goal is that working with Aizle once makes you permanently better at the thing you hired us for.

I should say this plainly → none of what you’ve just read works without a client brave enough to try it. Laura McGrath gave a new methodology real room to run, stayed open when the research challenged her assumptions, and trusted the work over the comfort of doing things the way everyone else does them. That kind of trust → putting your business behind someone else’s approach → isn’t something I take lightly.

“This has been one of the best decisions ever → and I have worked with a lot of branding people in the past, so I know how tricky it is to find someone who produces real, actionable results, without an excessive price tag.”

Laura McGrath, CEO Lola&Lykke

---

The Lola&Lykke case study is the most complete example I have of what happens when one senior strategic partner owns the entire thread → from the first research question to the last feedback prompt → with AI and behavioural science doing the heavy lifting and education built into every step.

If you’re entering a new market, repositioning a brand, or building a community → and you want the thinking AND the capability transfer → drop me a line at [email protected].