Early childhood education · Financial dashboard
TriMetrics ECE
An Iron Triangle financial dashboard for early childhood education — built from listening, finished on my own time.
The 55-second version
A purpose-built financial dashboard tracking the three metrics ECE operators need most — enrollment, revenue per child, and fee collection — in one place, with AI insights calibrated to whoever is looking.
Because I talked to the people who needed it, watched it get partially built, and finished it myself when it was shelved before it was done.
Four distinct personas — single-center owners, franchise operators, enterprise networks, and shared service providers — each with their own workspace.
The SSP workspace. It exists because someone told me directly what they needed and nobody else was listening closely enough to build it.
Background
Where it started
I work at a highly regarded, long-standing ECE software company. Trusted by the sector for decades, genuinely deep product — but software that's been around that long carries its own weight. Making changes is slow, and the company made a move I think was right: they acquired a newer online platform and started building toward a future where operators could run their programs from anywhere.
The problem is the online product and the legacy desktop have real gaps between them. Customers were being directed toward the online version — customers who needed things it wasn't fully built to do yet. I picked up a piece of the online product with the intention of closing some of those gaps. Specifically the financial ones.
We shipped two sides of the Iron Triangle. The third never made it out.
Research
What I heard before I built anything
Before I designed a single screen I spent time listening to the people this product was for. Not formal research — real conversations. Operators reaching out with frustrations. Sales calls where customers were realizing the product they'd bought didn't quite fit. Moments where someone described their day-to-day and I just paid attention.
Center owners
Exhausted by manual work. Enrollment in one place, invoicing in another, subsidy reimbursements falling through the cracks. They needed something that told them where they stood and what to do about it right now.
Franchise operators
Couldn't see their whole portfolio without building their own spreadsheets. The question they kept asking: which of my schools needs me today?
Enterprise networks
Needed a portfolio view at regional scale — a way to spot patterns before they became problems across networks spanning multiple geographies.
Shared service providers
Don't own the programs they support — they monitor health across programs they're responsible for without controlling. One SSP told me directly: "sometimes I don't know if the numbers are old." That sentence became a feature.
Validation
Holding my ground in the room
Before this went to customers, it went through a formal design review for sales and our shared service SME — people who'd been selling this feature for years off mockups that predated me. What my PM and I had built looked different, because it was built from what operators had actually told us, not from an old pitch.
The room pushed back hard. I didn't have a mentor there that day. I held my ground anyway, because the direction came from the people who'd actually use it — not a guess.
Customer validation settled it. Operators found it so useful I barely had to track adoption — it tracked itself. First time I trusted the research over the room, and it was the right call.
The gap
What got built — and what didn't
We shipped the enrollment tracking and fee collection pieces. And the SSPs who got access to even that partial version loved it. That told me everything — if two sides of the triangle were enough to generate that reaction, three sides done right would have been transformative for people who'd been promised a product that could meet their needs.
But the third side — revenue per child, the complete picture — never shipped. The project was deprioritized.
"I'd talked to the people. I knew what they needed. I watched some of them get sold on a product that couldn't fully deliver yet. So I finished it myself."
Design decisions
The rebuild
I rebuilt the whole thing from scratch, on my own time, as the version I always believed it could be.
The personas. Most tools treat "multiple sites" as one category. A franchise owner with five schools and an SSP supporting thirty independent programs are doing completely different jobs when they open a dashboard. The single-center owner gets everything on one screen with AI insights in plain language. The franchise owner gets a portfolio view where the question they kept asking — which of my schools needs me today — is answered the moment the page loads. The SSP workspace is the one I'm most proud of — the intervention queue, the program health table, the data reporting status column showing when each program last submitted data. None of that came from guessing. It came from someone describing their reality and me taking it seriously enough to build it.
Subsidy is woven in, not siloed. This decision came from a specific problem in the existing platform. When operators go to their finance section, they see cashflow from families. Subsidy lives on a separate tab. Most don't realize the two are disconnected — so when they pull numbers into their own spreadsheets, they're only capturing the first tab. Their totals are off and they don't know why.
In TriMetrics, subsidy never gets its own tab. The revenue per child figure shows the private pay and subsidy split directly. The fee collection rate flags when pending reimbursements would move the number if filed. If it's part of your financial reality, it shows up where you're already looking.
The onboarding routes people to the right place. A short questionnaire cross-references role with program count — an SSP managing a large portfolio gets a different workspace than a franchise owner with the same number of locations. The integration step connects to the tools operators already use. Real adoption requires meeting people where their data already lives.
What this is actually about
The SSPs who loved the partial version we shipped — they're the part that stays with me. They found value in something incomplete and told us clearly. That's evidence. Evidence that the need was real, the direction was right, and that the people who were sold a product that didn't fully fit them deserved something better.
I believe in human-centered design in a pretty straightforward way: you listen to what people are dealing with, you take it seriously enough to build something about it, and then you check whether it actually helped.
People told me what they needed. I built it.