Behind this prototype

The problem

• •Choosing a first sleep-away camp is an expensive, emotional decision made on marketing copy that all sounds the same.
• •It’s also scary and overwhelming: sending your kid away for the first time is a big leap, and there’s little plain-language education about what camp is or how it helps kids — which keeps it out of reach for many families.
• •The things that actually predict a kid’s experience — cabin culture, structure, social pace, homesickness support — aren’t listed anywhere, and no filter asks for them.
• •So families lean on word of mouth that fit someone else’s kid, or on “free” advisors paid by the camps they recommend.

Who it's for

• •First-time families. They have the least information and the most fear — and they matter most to camps: a well-matched first-timer can mean many summers of tuition.
• •The wider mission: more kids get to go to camp, not just a reshuffle of families already searching. That’s why the front page welcomes people who haven’t decided yet.

The name

A lanyard is the thing every camper makes and brings home — a few feet of gimp thread braided at rest hour, the small proof that they belonged somewhere. That is the product’s promise in one object: not just finding a camp, but finding the one where this kid belongs.

Strategy and positioning

• •Directory sites have breadth but no judgment — they filter on dates, activities, and price, and some sell placement.
• •Human advisors have judgment but no coverage — small rosters, paid by the camps on them.
• •Nobody has both, because the fit data exists nowhere. That’s the product: learn about the child in three minutes, get the camp’s side through onboarding, and show the reasoning on screen.
• •Against directories, we ask about the child, not the facilities. Against advisors: the whole market, the same flat fee for every camp, and we show the downsides a commissioned matchmaker leaves out.
• •Why now: AI makes the interview and the explanations cost seconds instead of a human hour.

How matching works

• •Hard limits filter first, in code — age, session, distance, budget, setting. The AI can’t override them.
• •AI ranks the remaining camps against your answers. Interests, social fit, and your goals weigh most; anything specific you wrote in your own words beats everything it touches.
• •The same AI writes each explanation, using only the camp’s structured data. Strength shows as a tier and a bar — never a percentage, which would be fake precision on sample data.
• •As real outcomes come in, a learned ranking model takes over the scoring. The target it learns from matters: completed, repeated summers — not clicks.

Business model, and why rankings stay honest

• •Camps pay a flat per-enrollment fee, identical for every camp, with one published rule: ranking is never for sale.
• •We link out freely — withholding information on a five-figure decision reads as steering, and steering is the thing we’re replacing. The intro is worth routing through us because it arrives with the family’s fit summary.

What I cut, and why

• •Real camp data — the fit attributes don’t exist publicly; scraping would invent them under real names. Labeled sample data is more honest. Browse the 30sample camps →
• •Reviews — they measure quality; the bet is that fit is the missing layer.
• •Accounts — the quiz is the personalization; a login would wall off the demo.
• •Scheduling plumbing — the request buttons prove the funnel; pipes prove nothing.
• •Kid and camp-director flows — three designed stubs mark where they go: the kid’s own landing page, the “Step 2: your camper weighs in” panel on the results page, and the director pitch page.

What I'd measure

• •Quiz completion — is three minutes of questions worth it? Target 80%+.
• •Conversations or tours on top-3 matches — hypothesis: 50% of completers either request a conversation or schedule a tour with at least one.
• •Agreement with independent human camp advisors on identical profiles — benchmark: 2 of top 3 overlap.
• •Kill signal: requests spread evenly across ranks means the explanations aren’t doing work.
• •North star: first-summer completion and second-summer return.

What this can and can't test

• •It tests whether explained ranking changes how a family shortlists.
• •It can’t test matching accuracy — that needs real summers.
• •One shortcut, named: distance matches region-to-region here; the real product uses actual geography.

What's next

• •Kid input (stub built) and camp onboarding (stub built) — verified fit attributes as the price of introductions.
• •Reference calls with current camp families, brokered by us.
• •Parent-set weights versus inferred weights, as an experiment. A post-summer outcome loop.