Context
When I was brought in, Koru Kids had a clear problem. Fewer families were verbally agreeing to hire a nanny than at the same point the previous year, and nobody inside the business was sure why.
Part of what made it hard to diagnose was how much had changed in the preceding twelve months. They'd automated a lot of the matching process, reshaped the team, and reworked their messaging. The side effect was that the people running the business were further from their customers than they used to be. They had theories about what was going wrong, but no direct line back to the families themselves.
Families were sampled from six points in the journey — some just signing up, some partway through matching, some who'd interviewed, some post-decision, some who'd hired, and some who'd given up. I wanted a spread across the arc rather than only the wins or only the drop-offs. I wanted the emotional texture of each stage as well as the operational facts, so interviews were semi-structured and went long where they needed to.
The team's six working theories
Before the research started, the team put forward six plausible explanations. Their instinct was to start designing interventions around all of them. I pushed back and asked to test them first.
Proved 3 of 6
-
Families waited to meet in person
An online-only decision felt too important for childcare.
-
Student timetables caused hesitation
Families needed certainty about hours before committing to a nanny still at university.
-
Families underestimated the timeline
Many assumed they had more time than the September rush actually allowed.
Disproved 3 of 6
-
Families were being too picky
Most had realistic expectations — they just weren't being met.
-
Cancelled interviews hurt reputation
Frustrating, but families saw them as individual-nanny issues, not platform ones.
-
The price increase was putting families off
Cost barely came up in any of the conversations.
Half of the team's intuitions were right. That's not a bad hit rate — but three plausible wrong answers can easily eat a quarter's roadmap. The research's first job was subtraction, not addition.
The family journey
Alongside the interviews, I mapped the end-to-end experience — from the first moment a family considered signing up, through to the first few weeks of a nanny actually working with them.
The map tracked what families were doing at each stage, what they were thinking, and how they were feeling. The emotional shape was more interesting than the operational one. Families felt relieved and excited early on, got anxious around the matching stage, recovered during interviews, and then dropped sharply in the post-interview moment where the real decisions were being made.
Where emotion runs ahead of the process
The lowest point sits between the interview and the decision — inside the part of the process the team's dashboards couldn't see.
Zooming in on the three stages that mattered
Most of the drop-off happened across three adjacent stages. I pulled them out into their own view so the team could see the shift from "hopeful" to "pressured" to "relieved" as a single, connected moment — not three separate operational steps.
"The kids seem comfortable with her."
"She's quite flexible with my work patterns."
"I'm not sure if she's experienced enough."
"Why do Koru Kids keep emailing me? I need time."
"Am I committing to 52 weeks now? How does pay work?"
"There's a risk a nanny won't turn up in September."
"What happens next?"
"Am I 100% committed?"
"Why haven't I heard back yet?"
Interview
"The kids seem comfortable with her."
"She's quite flexible with my work patterns."
"I'm not sure if she's experienced enough."
Train nannies on how to handle children of different ages. Make the vetting process more visible.
Post-interview
"Why do Koru Kids keep emailing me? I need time."
"Am I committing to 52 weeks now? How does pay work?"
"There's a risk a nanny won't turn up in September."
Slow down the chase. Let families set their own timelines. Explain what saying "yes" actually commits families to.
Hire
"What happens next?"
"Am I 100% committed?"
"Why haven't I heard back yet?"
Be clear about what happens after hire. Maintain contact with families before the job starts.
What was actually happening
The families who were stalling weren't stalling because they couldn't find a nanny they liked. They were stalling after finding one, in the gap between the interview and the decision — because three things were happening at once.
Families felt pressured to decide
They were pushing back against the automated chase sequence by doing nothing at all. The comms were designed to keep families engaged; they were reading as pushy at exactly the moment families needed space to think.
Families didn't know what "yes" meant
Ambiguity about the commitment — 52 weeks, pay as you go, what actually happened on day one — was enough to make them stall. The platform assumed families understood the answers; the research showed they didn't.
Families doubted nanny experience
Without visible evidence of the vetting process, families defaulted to scepticism and kept looking. They were told nannies were carefully selected, but not shown how.
None of these were on the team's hypothesis list. They were emotional friction points inside a part of the process the dashboards didn't cover — automated comms going out, families going quiet, and no visibility into why.
What I recommended
Rather than trying to fix everything, I focused on the post-interview moment where families were stalling, and on the ambiguity that was compounding the hesitation.
Slow down the chase.
Let families set their own timeline for deciding, and step back from automated nudges in the days after an interview. Reduce the sense of being pushed at the moment people most need space.
Rebuild the post-interview comms around the commitment.
Explain, concretely, what the nanny's first week will look like, what families are signing up for in terms of hours and duration, how pay works, and whether they can step back if things aren't working out.
Make the vetting process visible.
Show what Koru Kids looks for, what they screen out, and how training works. Give families something concrete to hold onto when they're weighing up the decision.
Outcome
Koru Kids acted on the post-interview communication piece quickly, rewriting the sequence to explain the process in more detail and give families clearer signposting about what came next. Slowing the automated chase came in a second wave. Vetting visibility was scoped for a later quarter.
Reflection
Koru Kids had built good dashboards, good automation, a strong matching algorithm. What they couldn't see was the five minutes after a family put the phone down from an interview — which turned out to be where the decision was actually being made.
The six hypotheses weren't wrong because the team was careless. They were wrong because they were generated from the parts of the system that were already visible, and the answer sat in a part of the system that wasn't. Research earns its place when it puts the invisible parts of a customer's experience back in front of the people making decisions about the product.