A 68% Checkout Abandonment at Seven Weeks
Pell & Co launched a direct-to-consumer skincare store seven weeks ago. It has had 22,000 sessions, a 4.2% add-to-cart rate, and 68% checkout abandonment. Average order value is $61 against a 65% gross margin. Shipping is a flat $7.50, and the free-shipping threshold is set at $75, which only 12% of orders reach. Mobile pages load in 4.8 seconds, and checkout accepts credit card and PayPal only, with account creation required before payment. Product pages show six studio photos, no reviews, and a 1,400-word ingredient essay above the fold. The bestselling serum has been out of stock for three weeks, yet ads still run to its product page, whose direct links now return a 404. Return rate is 9%.
1. Rank the three highest-impact fixes and estimate what each is worth, showing your arithmetic.
Read the worked answer
Size the funnel first: 22,000 sessions × 4.2% add to cart = 924 carts; at a 32% completion rate that is about 296 orders and roughly $18,000 of revenue. The abandonment pool is 924 − 296 = 628 carts per period.
Fix 1: stop the bleeding on the out-of-stock serum. It is the worst value-per-hour item on the list - ads are paying for clicks that land on a 404. Redirect the URL to a category page, replace the ad creative, and put a back-in-stock capture on the product page. Cost to fix is near zero; the loss is the full ad spend plus every sale that would have happened.
Fix 2: checkout friction. Forced account creation and a card-plus-PayPal-only wallet set are classic abandonment drivers. Add guest checkout and Apple/Google Pay, and show shipping cost on the cart page. Recovering even 10% of 628 abandoned carts is 63 orders × $61 ≈ $3,800 per period at essentially no media cost.
Fix 3: mobile speed from 4.8 seconds. Treat any improvement as something to validate rather than a guaranteed percentage, but a slower-than-3-second mobile store is losing sessions before the add-to-cart event ever fires. Reviews and moving the essay below the fold come after these, because they affect conversion rate, not abandonment.
2. Should Pell & Co lower, raise, or remove the $75 free-shipping threshold? Show the margin logic.
Read the worked answer
At an AOV of $61 with 65% margin, gross profit per order is $39.65. If shipping is subsidized on that order, the $7.50 comes straight out of margin, leaving $32.15 - a 19% cut in gross profit per order. To hold total gross profit flat you would need about 23% more orders, which a shipping offer alone will not deliver.
Removing the threshold therefore subsidizes every order and is the most expensive option. Keeping it at $75 means only the 12% of customers who already reach it get the subsidy, which is cheap but barely changes behaviour. The interesting question is the gap: a customer at $61 must add $14 to qualify. That $14 of extra product earns $14 × 0.65 = $9.10 gross profit against the $7.50 shipping cost, so the threshold is only accretive if it is set where the required add-on clears $7.50 ÷ 0.65 = $11.54.
Recommendation: test a threshold at $70-75 with a progress bar showing how close the cart is, and measure AOV lift against the shipping cost. Do not remove it. If AOV lift does not clear $11.54, ship the cost into the displayed price instead.
3. Design the checkout investigation and the test plan that follows it.
Read the worked answer
Instrument before you test. Record every checkout step - cart, contact information, shipping method, payment - and read drop-off by step and by device. A 68% figure alone cannot distinguish shipping-cost shock (a spike at the shipping step) from forced account creation (a spike at the contact step), and the fixes are entirely different.
Add two qualitative inputs: an exit-intent survey asking why the shopper is leaving, and session recordings on mobile checkout. Segment everything by new versus returning customers, since returning customers will tolerate an account they already have.
Then test in sequence, one variable at a time. Start with guest checkout, because it is the most likely culprit given account creation is mandatory and the fix is configuration rather than design. Baseline completion is 32%; to detect a 3-percentage-point absolute lift at reasonable power you need traffic on the order of tens of thousands of checkout starts, so run it long enough rather than calling it in a week. Primary metric is checkout completion; guardrails are AOV, refund rate, and the 9% return rate, which should not rise.