Mapping the Loop: A Systems Audit of a Sleep-Brand Retention Engine
Objective: Given a quarter of funnel and loop data from a DTC mattress brand, apply the lesson's systems-audit steps to classify loops, flag the delay that's hiding the real cause, and name the single constraint.
You're the growth analyst at Casper. Leadership is confused: referral-driven signups have been flat for two quarters despite review volume climbing every month. You've been handed the loop map and funnel numbers to explain why.
Classify the loops as reinforcing or balancing, mark where the multi-month delay sits, and name the constraint before recommending a single fix.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free, handles the pivot and chart work without new tooling
Free, connects directly to the funnel export for repeatable quarterly audits
The process
2 steps
Step 01 of 02
The lesson splits loops into reinforcing (they compound, like happy customers referring more customers) and balancing (they self-limit, like rising CAC capping growth).
The data shows: verified reviews grow every month, review-driven signups grow with them, but 'signups per new review' has been declining each of the last 3 months. Which loop is actually running, and is it healthy?
Procedure
- Import loop-data.csv and chart reviews (x) against review-driven signups (y) by month
- Compute signups-per-review for each month
- Flag the month the ratio started declining
Month Reviews Review-driven signups Signups/review Jan 1,200 340 0.283 Feb 1,450 372 0.257 Mar 1,690 379 0.224
Healthy
A reinforcing loop where signups-per-review holds steady or grows as review volume climbs, proof the loop is still compounding.
Unhealthy
A reinforcing loop with a steadily declining ratio, the loop is still positive but weakening, most often because the review pool is diluting with lower-intent or lower-quality reviews.
What this means
Reviews are still reinforcing signups in raw terms, but the declining per-review yield means something upstream, likely review quality or targeting, is quietly degrading the loop before it shows up as a total decline.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Signups-per-review has fallen 3 months straight | Audit the last 90 days of new reviews for star rating and specificity before touching ad spend | 30 min |
Step 02 of 02
The lesson's 4-step audit ends by locating the one constraint governing total output, after mapping stocks/flows, finding loops, and marking delays.
Funnel data shows: site visits up 22% this quarter, add-to-cart flat, trial starts down 4%. The referral loop's delay is 4-6 months (a review posted today reflects a mattress bought 4-6 months ago). Where is the constraint, and what would adding more ad spend do to it?
Procedure
- Pull visits, add-to-cart, and trial-start counts for the last 2 quarters
- Compute stage conversion rate for each transition
- Identify which transition's rate dropped the most
Stage This Q Last Q Delta conversion Visit to Add-to-cart 3.1% 3.4% -0.3pt Add-to-cart to Trial 41% 52% -11pt
Healthy
Ad spend increases feed a funnel where every downstream stage's conversion rate holds, more traffic becomes more trials proportionally.
Unhealthy
Ad spend increases feed a funnel where a downstream stage (here, add-to-cart to trial) is dropping in conversion, meaning new traffic backs up behind that stage instead of converting.
What this means
The constraint is add-to-cart to trial, down 11 points, not top-of-funnel traffic, which is actually up. More ad spend right now buys more visitors stuck at the same broken stage, not more trials.
So what do I do about it?
| Symptom | Action | Effort |
|---|---|---|
| Trial starts fell while visits rose | Investigate the trial-start step (checkout friction, financing options, delivery date shown) before approving next quarter's media budget increase | half day |
Final deliverable
A one-page systems audit memo: the loop classified with its current health, the constraint identified with supporting numbers, and one recommended action.
See a reference example
Allbirds Q2 systems audit (excerpt) LOOP: Referral loop (reinforcing, weakening) Signups/review down from 0.28 to 0.22 over 3 months Likely cause: review quality dilution from bundle promo CONSTRAINT: Landing page to email-capture (41% -> 29%) NOT ad spend, NOT traffic RECOMMENDATION: Fix the email-capture form before increasing paid spend.
Success criteria
You're done when you can:
- Correctly classifies the loop and its current trend, not just its type
- Identifies the actual constraint stage with numbers, not a guess
- Recommendation targets the constraint, not the symptom