Once your upsells are live, the next step is improving them. The best-performing merchants review results regularly, change one thing at a time, and use data to decide what to keep.
Use this guide to benchmark your results, run a weekly optimization cycle, test improvements manually, and decide when to expand to another page or placement.
Start by checking the metrics that show whether your upsells are being seen, accepted, and increasing order value.
Metric | What it tells you | How to use it |
Conversion Rate | How often customers accept an upsell after seeing it. | Use it to judge product relevance and offer strength. |
Edit to Upsell Rate | How often edit sessions result in a paid upsell. | Use it as a health metric across your self-service editing flow. |
AOV lift | How much upsells increase average order value. | Use it to confirm revenue impact, not just accepts. |
Top Upsell Products | Which products drive the most views, accepts, and revenue. | Use it to double down on winners and replace weak products. |
Surface Coverage | Which pages and placements are contributing revenue. | Use it to find missing or underperforming placements. |
ℹ️ Note: Conversion Rate and Edit to Upsell Rate measure different things. Review both before deciding whether a strategy is working.
Benchmarks help you decide what to fix first. Treat them as a guide, not a guarantee, because product type, placement, traffic, and price all affect performance.
Performance tier | Conversion Rate | Edit to Upsell Rate | What to do next |
Needs tuning | Below 2% | Below 5% | Check installation, filters, product relevance, and pricing. |
On track | 2 to 5% | 5 to 10% | Keep improving product selection and test one new idea. |
High performing | Above 5% | Above 10% | Protect what works and test advanced targeting. |
Give a new strategy at least 7 to 14 days before judging results. If traffic is low, wait until the strategy has enough views to show a clear pattern.
Small, consistent improvements usually beat large one-time changes. Review your upsells once a week and change one variable at a time.
Check your trend. Compare Conversion Rate, Edit to Upsell Rate, AOV lift, and revenue against the previous week.
Review Top Upsell Products. Keep products with strong accepts and revenue. Replace products with high views and low accepts.
Review AOV segments. If below-AOV customers don't convert, test lower-priced products. If above-AOV customers convert well, test premium products or bundles.
Make one change. Change the product, discount, headline, placement, or filters. Don't change everything at once.
Record what changed. Use Strategy Logs to confirm what was saved and when.
Order Editing doesn't currently include built-in split testing. You can still improve performance by testing one change at a time and giving each change enough traffic before judging it.
What to test | Current version | Next version |
Product selection | Current bestseller | More relevant complementary product |
Discount | No discount | Small discount, such as 5 to 10% |
Copy | Generic product headline | Clear benefit or completion message |
Filters | Show to all customers | Filter by product, collection, or order value |
Keep a short note of what changed and when. Compare the same metrics before and after the change, especially Conversion Rate, AOV lift, and Top Upsell Products.
If an upsell underperforms, product selection is usually the best first thing to fix. A better product match often beats a bigger discount.
Problem | Likely cause | Try this |
High views, low accepts | The product isn't relevant enough. | Use Complementary Products, Metafields, or a more specific Collection. |
Low views | The strategy may not be matching many customers. | Review filters and test broader matching. |
Accepts but low revenue | The product price may be too low. | Test a higher-value add-on or bundle on the Post-Purchase Page. |
Revenue but weak margin | The discount may be too generous. | Test a smaller discount or stronger value-based copy. |
Once one strategy is performing, expand to another page or placement. Each page catches customers at a different moment.
If you're using | Try next | Why |
Checkout Page only | Thank You Page or Order Status Page | These pages are available on more plans and don't interrupt checkout. |
Thank You Page only | Order Status Page | Customers may return to view tracking and still have time to add products. |
Post-purchase placements only | Checkout Page or Free Shipping Upsell | You can catch customers before payment with low-friction add-ons. |
Check Surface Coverage after expanding. If one page gets views but little revenue, improve the products before adding more placements.
A Fallback Strategy gives Order Editing something to show when no targeted strategy matches. It protects coverage and keeps broad-appeal products available.
Use products with wide relevance, strong margins, and simple decision-making. Bestsellers, accessories, gift cards, care kits, and low-priced consumables often work well.
💡 Tip: A Fallback Strategy shouldn't be your most complex setup. Keep it simple, broad, and easy for most customers to understand.
Showing too many products. Start with a small set of strong products. Too many choices can reduce accepts.
Discounting before fixing relevance. If the product doesn't fit the order, a deeper discount may only reduce margin.
Changing too much at once. If you change the product, copy, discount, and filters together, you won't know what worked.
Ignoring mobile preview. Most customers will see a compact layout. Preview mobile before setting a strategy Active.
Forgetting Draft status. A strategy won't show to customers until it's Active.
