Where this ends. What it returns.

Strategic acquisition (primary, 55% probability, Y5 to Y7, 15 to 25x). Tekku becomes the kid-build surface inside a larger platform. Primary candidates are the edu-aligned platforms (Duolingo, Khan Academy parallel, Code.org, Roblox, Common Sense Networks) and the horizontal AI platforms adding a vertical youth channel (Anthropic, OpenAI, Google, Apple, Microsoft, Disney). The concept graph is the second-order asset the acquirer buys alongside the subscriber base. Valuation sits at the revenue and EBITDA multiples the IPEV table above anchors.

Secondary in growth round (bridge, lower probability, Y4 to Y6). A Series B or C round at $50M+ sizing with a secondary tender component. Not a full exit, but a liquidity event for seed participants who want to realize part of the position while the company continues to compound. Precedent in the category is common once ARR is above $25M and retention is durable.

IPO (long-arc, 15% probability, Y7 to Y10, 20 to 40x). Public markets path once ARR clears $250M and the Workshop motion prints a predictable enterprise revenue stream alongside the consumer subscription. Duolingo at IPO (18x revenue, traded up) is the pattern. The public-markets comp set for AI-native consumer edtech will be mature by 2029 to 2030, which is where our IPO window sits.

Seed-stage SAFEs convert into priced-seed equity at the Series A, which in our timeline is Month 18 with a target close Q4 2027. The priced-seed equity then rides through Series B (Y3 to Y4) and Y5 to Y7 exit. Primary liquidity event is the exit; the J-curve is visible on paper through the two intervening priced rounds, where mark-ups are unrealized but auditable.

Secondary options emerge at the Series B (Y3 to Y4). Most growth-stage rounds of that sizing allow a small secondary tender as part of the primary. This gives seed participants a partial-liquidity path before the full exit if they want it. The company does not push secondary; it is an investor-side option, not a founder-side plan.

The $18M post-money cap means a $25K check converts at cap into the equivalent of buying at an $18M post-money priced round (see the Offer page worked examples). At the base-case exit EV of $2.6B with standard dilution (Series A ~20%, Series B ~18%), the compounded gross return before fees is in the 3 to 4x band for tier-1 SAFEs. At the upside IRR, 6 to 10x. At the downside, a venture-grade return that clears cost of capital.

Brand. Parent-trust in a kid-facing AI product is the scarcest commodity in the category. Every moderation incident in the cohort of kid-AI incumbents compounds the premium on a product that has earned parent trust cleanly. The Kaze creative doctrine (seven laws of living platforms) is the internal spine; the external visible output is a weekly email every parent actually reads. Brand is worth roughly 10 to 15% of the multiple at exit.

Data moat. The concept graph is the only dataset in the world that describes how children learn to build with AI. It is labeled every session, detected rather than assigned, and it compounds faster the more kids ship. Acquirers price this asset separately from the subscriber base. Data moat is worth another 10 to 15% of the multiple.

Margin profile. A 68% gross margin and a 30% EBITDA margin at scale put Tekku in the same multiple cohort as premium SaaS, not the compressed cohort of consumer edtech. AI inference cost curves accrete margin over time, which supports multiple expansion rather than compression as the company matures.

Category leadership. The AI-native kid-build category is forming now, with the compliance window running 2026 to 2028. A product that is compliance-first from day one and category-defining by Y2 commands the category-leader premium. The alternative is retrofit incumbents climbing back uphill into a market that already named the winner. Category leadership is the largest single premium driver, worth 15 to 25% of the multiple at exit.