Rimu Capital closed Fund I in June 2023 after eighteen months of conversations with LPs, founders, and educators across New Zealand and Australia. What I didn't fully anticipate, even after five years in education policy at the Ministry of Education and another five building product inside a school-facing LMS, was how much the act of forming a thesis would reshape what I thought I understood about the sector. Investing forces a different kind of rigour than operating does. When you're inside a product team, you're optimising for what users need right now. When you're backing companies, you have to take a position on what the market will reward in seven years.
The first lesson was about buyer complexity. New Zealand schools sit in one of the more unusual procurement environments in the OECD. Principals have genuine spending autonomy — far more than their counterparts in the UK or Australia's state systems — but that autonomy is bounded by the Ministry's digital infrastructure decisions, the resourcing agreements negotiated through school property clusters, and the practical reality that most schools run their ICT decisions through a two-person board of trustees committee. What looks like a fragmented market (thousands of autonomous buyers) turns out to be a market where six or seven decisions — a handful of regional coordinators, the NZCER relationships, the Enabling e-Learning platforms — shape what most schools actually adopt. The implication for founders is that direct-to-school sales and channel-through-cluster are not interchangeable strategies. We backed a team in our first year whose go-to-market assumed direct-sale economics; reality was that they needed a cluster-led channel and the unit economics look completely different. We got that founder through to the right model, but it cost eighteen months.
The second lesson was about the difference between what teachers say they want and what they actually adopt. Every founder who has done classroom research comes back with the same finding: teachers say they want tools that reduce administrative burden and help them personalise instruction. When those tools arrive, adoption is deeply uneven. The gap is almost never about feature quality. It's about whether the tool fits inside the existing rhythms of a school day — the way a lesson follows a registration period follows a bell schedule follows a reporting period. Products that require teachers to change when they check in on students, or that create new data-entry workflows before the assessment machine yields any output, fail even when the underlying pedagogy is excellent. We now ask founders to map, explicitly, where in the school-day rhythm their tool inserts itself. If the answer is "teachers will use it during planning time," we ask what planning time looks like in the specific schools they've piloted in — because planning time in a rural NZ primary is structurally different from planning time in an Auckland secondary.
We are not saying NZ schools are unusually difficult buyers. In many respects the opposite is true — NZ principals are exceptionally willing to pilot, the Ministry has created genuine on-ramps for ed-tech through initiatives like the Digital Technologies curriculum reform, and there is a culture of practitioner-led innovation that you don't find at the same density in most markets. What we are saying is that willingness to pilot is not the same thing as willingness to pay at scale, and the distance between those two is where a lot of early-stage EdTech companies stall. Our diligence now includes a specific question: has this company converted at least one pilot school into a paying renewal at institutional budget (not innovation-line) rate? If not, we treat the product as pre-commercial regardless of the pilot count.
The thing I would do differently is spend less time in the first six months trying to construct a comprehensive sector map and more time sitting with founders who were already building. The sector map was useful — it helped us articulate the thesis to LPs, and it gave James a framework for technical diligence. But the most important signals we found in year one came from conversations with people who were three or four years into building in the Australasian market and had the specific texture of failure and near-failure that no desk research produces. Aroha's background in impact fund analysis gave us a template for that kind of qualitative diligence. The goal for year two is to run it more systematically, and to extend the founder network into Southeast Asia, where the structural dynamics of formal schooling have interesting parallels with what we see in NZ — large public systems, underfunded assessment infrastructure, and a growing middle class willing to invest in supplementary education tools.