Coffee Kit economics and experiment design
12 November 2026. All numerical assumptions below are planning scenarios, not measured forecasts.
Revenue and contribution
Gross ARR = active paying stores × realized monthly ARPU × 12. At $99, 500 stores yield $594,000 and 842 stores yield $1,000,296. Realized ARPU should include discounts, refunds and plan mix rather than merely the advertised top plan.
The companion CSV models $49/$79/$99/$149/$199 monthly plans at 10/30/60 minutes support per shop-month, $30 loaded support hourly cost, $5 hosting/AI allocation and a 5% platform/payment reserve. The reserve is deliberately a scenario, not a statement of Shopify's fee schedule. Replace all with actual invoices and time logs before a budget decision.
At $99 the resulting contribution is $84.05/$74.05/$59.05; this excludes fixed engineering, design, acquisition and tax. A $300 acquisition cost would repay in 3.57/4.05/5.08 months under those respective scenarios, before churn and onboarding. Two hours of assisted onboarding adds $60 at the same labor rate.
The illustrative $15,000 fixed monthly cost includes only costs not already attributed as support or per-shop hosting. If support is an existing salaried role, keep two views: cash operating budget using that salary once, and economic contribution allocating their time. Do not subtract the same salary in both. Customer subscription-provider, email and loyalty fees belong in the merchant's total stack cost, even if they are not Coffee Kit COGS.
The support ceiling
At 500 stores, 10 minutes/store/month consumes 83.3 hours; 30 minutes consumes 250 hours; 60 minutes consumes 500 hours. A support person with 80 hours/month available after theme support, prospect work and admin can handle 480/160/80 stores respectively. Even the optimistic case leaves little onboarding capacity at 500 stores.
Measure support by workflow: initial mapping, theme installation, catalog correction, provider issue, renewal/substitution, billing and general education. Automation should reduce repeated explanations and detect errors; it cannot turn unlimited custom integration into a three-person business.
Proposed service boundaries: supported theme/provider matrix, guided import, one documented installation path, response hours, incident escalation, and paid agency services for bespoke design. Avoid an enterprise concierge promise in a $99 self-serve plan.
Acquisition and churn scenarios
With constant new paid stores A per month and monthly logo churn c, active stores after month m are A × (1 − (1 − c)^m) / c, starting at zero. These assumptions ignore seasonality, delayed activation, cohorts and expanding ARPU.
At 30 new paid stores per month and 3% monthly churn, the long-run ceiling is 1,000 stores; reaching 842 takes about 61 months. At 30/month and 5% churn, the ceiling is 600: $1M ARR at $99 is unreachable without better retention, acquisition or ARPU. At 30/month and 2% churn, reaching 842 takes about 41 months. These are arithmetic scenarios, not predicted churn.
At 10% qualified-demo-to-paid conversion, 30 new paying shops requires 300 qualified demos per month. That is unsuitable for a founder-led one-to-one funnel with this team. Scale would require better conversion, self-serve activation and/or agency/provider distribution. A premium plan does not remove the need for scalable distribution.
Merchant value test
Use incremental contribution rather than attributed revenue as the primary benefit. Example assumptions: $40 incremental order, 45% contribution after coffee, fulfillment, payment and incremental discounts → $18/order. A $99 app needs roughly six genuinely incremental orders/month to cover its fee; a combined $199 stack needs roughly twelve. This does not prove the app will create those orders.
A subscription can move an existing buyer's orders into a different billing route without increasing profit. Track retention, order frequency, contribution, discount cost and cancellations versus the best available baseline. Time saved is a second value component, measured with actual repeated catalog tasks. Avoid counting both the same saved work and replaced agency fees twice.
For smaller stores, do not try to justify $99 with a few minutes of data entry. Offer a constrained data/showcase plan only if the margin and support burden work, or a one-time onboarding service through partners. For merchants with existing subscription revenue, test whether the improved enrollment flow and reduced rotation work justify the additional app fee while keeping their engine.
Pilot design
Recruit 8–12 qualified stores across Coffeetale/Krown and at least one external supported theme. Favor shops with recurring online bean sales and a real subscription or next launch scheduled. Include a stable-blend control and seasonal-lot shops; do not assume every coffee shop has the same need.
Observe the existing process before showing the solution. Ask for the next actual product sheet, a recent product launch, current subscription workflow, examples of sold-out substitutions, support messages and current app invoices. Record baseline task time, failure modes, support minutes and contribution assumptions. Then run the same task with the prototype or supported pilot.
Proposed decision gates, selected for this project rather than universal startup benchmarks: - Technical: one provider can support the promised lifecycle with documented permissions and affordable fees; stock mapping, cancellation and renewal exceptions behave correctly in a development store. - Commercial: at least five qualified merchants accept the actual proposed ongoing price and defined supported scope; record reasons for all declines. - Activation: at least four launch a useful experience with their own catalog; founder effort is measured, not hidden. - Repeat: at least four complete a second catalog/club change; at least three renew after the initial paid period with support trending toward the target. - Value: merchants identify observed saved work or useful subscription outcomes, and no critical product/payment/stock corruption occurs. - Reach: at least two paid pilots come from outside the existing Krown relationship before claiming independent distribution.
Small samples give directional evidence, not statistically established conversion uplift. Do not stop solely because an arbitrary signup count was missed; diagnose reach, qualification, offer, activation and value separately.
Cheapest falsification tests
- Show a merchant's next real coffee in the actual Coffeetale card and builder, with the planned price and existing provider fee together.
- Ask them to perform the second seasonal product change without the founder operating the UI.
- Test one expiring-lot subscription change with the candidate provider in a dev store before building a broad configuration system.
- Compare against native Shopify/Sidekick plus the Coffeetale guide for the same task.
- Ask whether they would keep paying after switching themes. This tests the app's independent value.
Native platform control
Sidekick-generated apps can already use Admin data including products, metaobjects and discounts. Their documented generated-app surface remains inside admin, so compare actual tasks rather than claiming Sidekick does nothing or replaces the whole proposed product. Shopify Sidekick apps.
Canvas currently supports Shopify-developed/custom themes and cannot add/configure app blocks. These present limits are not a durable moat. Merchant-owned facts and reusable experience configuration should survive a theme change even if a new presentation adapter is necessary. Canvas requirements.
Actual Shopify fee sensitivity
The official schedule observed on 12 November 2026 gives eligible developers 0% revenue share on the first $1M in lifetime gross app revenue counted from 1 January 2025, then 15%; the threshold is aggregated across associated developer accounts and does not reset annually. Billing also incurs 2.9% processing, with applicable taxes/regional fees. Theme Store and referral income are excluded from that app threshold. Confirm Krown's actual app-revenue position before budgeting. Shopify revenue share.
At $99 with $5 hosting/AI and 30 minutes support at $30/hour, using 2.9% gives $76.13 contribution; using a simple additive 17.9% fee scenario gives $61.28 before tax/regional adjustments. The latter requires 245 shops to cover $15,000 of otherwise-unallocated monthly fixed costs. The original 5% reserve CSV remains a deliberately labeled early-stage scenario, not a mature-scale margin forecast.
Source: economics-validation.md The complete Markdown handoff and private datasets remain in the local project. This web edition includes selected supporting visuals and scenario downloads.