Discounts feel like a lever. They are a leak.
Every Shopify brand we meet has a discount problem. The codes started as a quarterly push and turned into the default. Customers learned the cadence. Conversion rate is up, contribution margin is down, and nobody on the team can remember the last full-price week.
The fix is not a clever code. It is a different surface. The cart drawer is the highest-leverage place in the funnel to lift AOV because intent is already proven. The customer has put something in the bag. The next click is yours to influence.
Pattern one: the threshold reward
Free shipping is the most studied AOV mechanic on the internet, and most stores still implement it wrong. The win condition is not a banner. It is a progress bar wired to the actual cart subtotal, refreshing in real time, with a copy variant for under-threshold and a celebration variant for over-threshold.
We see 6 to 14 percent AOV lift on this single pattern when the threshold is set 18 to 22 percent above current AOV. Set it too high and customers ignore it. Set it too low and you are giving away shipping you would have eaten anyway.
Pattern two: the recommended add
The slot under the cart subtotal is the most valuable real estate on your store. It is the last surface before checkout, and it is where most stores leave money on the table by showing nothing, or worse, showing the same product the customer already added.
Train it on co-purchase data from the last 90 days. Filter to items under 35 percent of average cart subtotal. Show one product, not three. Single-decision surfaces convert 2 to 3x better than carousels because they remove the choice tax.
Pattern three: the bundle nudge
If a customer has a hero SKU in their cart and you sell a kit that includes it, the cart drawer is where you say so. Not on the PDP, where the customer has already chosen. In the drawer, where the next click is checkout or close.
The copy matters. Compare to the a la carte total, not a percentage off. Customers do percentages poorly and they do dollar deltas well. A 47 dollar savings beats a 22 percent savings even when the dollar figure is smaller.
What to measure, what to ignore
Track AOV by cart subtotal cohort, not blended. A blended AOV number hides the truth. The cohort that hit your threshold reward is the only one whose behavior changed. Looking at the blended line tells you nothing about whether the mechanic worked.
Ignore opens, ignore clicks on the drawer chrome, ignore the time-to-checkout vanity numbers some apps surface. The only number that matters is incremental contribution margin per session, post the cart drawer event. If that is up and your discount rate is flat, the pattern is working.