Build the Wheel: Mapping a Subscription Business's Attract-Engage-Delight Loop
Objective: Given a subscription business's current channels, sales process, and post-purchase experience, build a complete flywheel map (Attract/Engage/Delight, one friction point and one force multiplier per underbuilt arc) that a growth team could act on.
You're a growth strategist advising Warby Parker's contact-lens subscription line, still under 12 months old, on how to plan its flywheel instead of a funnel.
Map the current state of all three arcs with real inputs, then add exactly one force multiplier proposal for the arc with the fewest touchpoints so the deliverable is prioritized, not just descriptive.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free tier supports a single unlimited board, enough for this exercise
Free, quick single-paragraph deliverable
The process
2 steps
Step 01 of 02
The lesson's Step 1 asks: what content or channels bring people to you (Attract), how do people become paying customers (Engage), what happens after the sale (Delight).
Given the subscription line's current channels -- Instagram, a referral link in the shipping box, and email -- sort each into its correct arc and flag the arc with the fewest inputs.
Procedure
- Create Attract / Engage / Delight columns
- Place every known channel or touchpoint into its correct column
- Circle the column with the fewest stickies -- that's the underbuilt arc
ATTRACT: Instagram ads, organic try-on content ENGAGE: online style quiz, checkout flow DELIGHT: (empty) Delight has zero mapped touchpoints.
Healthy
All three arcs have at least one real touchpoint mapped.
Unhealthy
One arc is empty, meaning the business is planning a funnel, not a wheel.
What this means
An empty Delight column means referrals and repeat purchases are left entirely to chance.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Delight column has zero touchpoints | Add a first-order check-in email and a referral prompt in the shipping box before spending more on Attract | 30 min |
Step 02 of 02
The lesson's Step 3 lists referral programs, user-generated content, integrations, and community as force multipliers.
Given the empty Delight arc from Step 1, which force multiplier from the lesson's list fits a contact-lens subscription business best, and what specific mechanic would you launch?
Procedure
- Pick one force multiplier type from the lesson's list
- Specify the exact mechanic, not just the category
- State which arc it feeds back into
Force multiplier: referral program. Mechanic: 'Give a friend their first month free, get $10 off your next order' link included in every shipping box. Feeds back into: Attract (new signups) and Delight (existing customers feel rewarded).
Healthy
Mechanic is specific enough to hand to an engineer or ops team.
Unhealthy
Proposal stays at the category level ('do referrals') with no concrete mechanic.
What this means
A force multiplier only works once it's a specific, shippable mechanic, not a category label.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Force multiplier proposal is too vague to build | Rewrite with the exact incentive amount and where it appears in the customer journey | 30 min |
Final deliverable
A 3-arc flywheel map (Miro board or equivalent) plus a one-paragraph force-multiplier proposal for the underbuilt arc.
See a reference example
Instacart same-day grocery flywheel map (excerpt) ATTRACT: paid search, in-app restaurant cross-promotion ENGAGE: one-click reorder, saved carts DELIGHT: (empty before this exercise) Force multiplier proposal: 'refer a neighbor, both get $15 off' prompt shown after a 5-star delivery rating. Feeds Attract (new signups from existing dense-delivery zones) and Delight (existing customer feels rewarded right after a good experience).
Success criteria
You're done when you can:
- All three arcs are populated with at least one real touchpoint
- The underbuilt arc is correctly identified before proposing a fix
- The force-multiplier proposal specifies an exact mechanic, not just a category