Find the Leaks: Teardown of a Broken Cart-Recovery Sequence
Objective: Given a synthetic 3-email cart-recovery sequence for Chewy, find every defect that violates the lesson's timing, discount, and personalization rules, and separate real defects from plausible-looking non-issues.
You're a contractor auditing Chewy's Autoship cart-recovery flow before a subscription push. The existing sequence was built two years ago and nobody has touched it since.
Read all three emails in the specimen sequence, flag every rule violation against the lesson's timing and discount-sequencing framework, and don't flag the two lines that are actually correct.
Before you start
What you'll need
Free path (everything below is enough to finish)
Free tier includes the flow builder view needed to check send-delay logic
Free, no account friction, easy to share as a deliverable
Paid upgrades (optional, faster/deeper)
Paid plans unlock multi-branch flow logic beyond the free tier's basic triggers
The process
Specimens to review
List every defect in this sequence against the lesson's rules on timing, discount sequencing, product data, and scarcity claims. Rate each by severity and note what a healthy version would do instead.
EMAIL 1 (sent 5 days after cart abandonment) Subject: "Don't miss out!!!" Body: "Hi there! We noticed you left something behind. Use code SAVE15 for 15% off your order today only!" [No product image, no item name, no price shown] EMAIL 2 (sent 6 days after abandonment, 24 hrs after Email 1) Subject: "Still deciding?" Body: "Thousands of pet parents trust us. Check out our 4.8-star reviews! Free shipping on orders over $49." [Includes a photo carousel of 3 customer review screenshots] EMAIL 3 (sent 9 days after abandonment) Subject: "ONLY 2 LEFT IN STOCK, ACT NOW!!!" Body: "This item is almost gone. Grab it before it's too late. Extra 10% off with code LASTCHANCE, expires in 24 hours." [Inventory system shows 340 units in stock for this SKU]
Specimen: synthetic, realistic
Final deliverable
A defect log (spreadsheet) listing each flaw found, its severity, the lesson rule it violates, and the one-line fix, plus a note on which two flagged items were actually correct.
See a reference example
Rent the Runway cart sequence audit (excerpt) CRITICAL: Email 1 sent 4 days late -> move to 45-min trigger CRITICAL: Email 1 leads with 20% off -> remove discount, add product image MODERATE: Email 3 scarcity claim unverified -> pull live inventory count before sending CORRECT (not a defect): Email 2 review carousel + free shipping mention
Success criteria
You're done when you can:
- Flags all 4 critical defects with the correct lesson reference
- Does not flag either of the 2 correct (non-defect) sequence elements
- Assigns a severity to each defect found