Bundles, contextual cross-sells and a properly calibrated spend threshold may lift average order value when they’re tested before being scaled. The ceiling moves with category and price point, so skincare and furniture retailers will see different ceilings. Measurement only means anything if you segment it: new versus returning customers, channel, device.
Key Takeaways
The most effective way to increase average order value is by testing and applying targeted tactics in order, while carefully measuring segment-specific results and margins.
| Point | Details |
|---|---|
| Test in sequence | Implement one AOV tactic at a time, then measure its impact before adding another. |
| Segment measurements | Break down data by customer type, channel, device, and category to accurately assess changes. |
| Use clear bundles | Show individual prices alongside bundle offers to make savings obvious and encourage purchases. |
| Optimize placement and info | Place suggestions on pages where customers decide or are about to buy, and include key details for conversion. |
| Leverage Thrive-agency | We provide comprehensive support, combining optimization, creative production, and strategic campaigns. |
Table of Contents
- The highest-leverage AOV tactics, ranked
- Calculate your baseline AOV and segment it so you can measure changes
- Product bundling: when to bundle, bundle types and how to price them
- Upsell and cross-sell best practice: placement, messaging and product attributes
- Spend thresholds and incentives: how to pick the right free-shipping or gift cutoff
- Personalisation and AI-driven recommendations: what to personalise and where
- Post-purchase offers and checkout protection
- Measure, test and iterate: a practical test plan
- Implementation checklist and tool categories to deploy quickly
- How we think about AOV lifts
- Want a hand running this properly
- FAQ
- Sources
The highest-leverage AOV tactics, ranked
Most stores try everything at once and can’t tell what worked. Test in this order instead.
- Bundles: best when you sell complementary SKUs (a razor and blades, a camera and a memory card); the uplift comes from making the bundle maths obvious, not from hiding a discount.
- Contextual cross-sells and upsells: best on product and cart pages where the recommendation is genuinely related to what’s already in the basket.
- Spend thresholds: best when a meaningful share of orders already sit just below an achievable cutoff; useless if nobody’s close.
- Post-purchase offers: best for low-friction, low-cost add-ons shown after the card details are already in, not before.
- Personalisation: best where you have enough purchase history or session data to beat a generic “you might also like” carousel.
- Loyalty tweaks: best for repeat buyers where a points threshold nudges the next order up rather than just rewarding the one just placed.
The failure mode is the same across all six: push too hard and conversion drops faster than AOV rises. A threshold set too high just gets ignored. An upsell shown at every single touchpoint trains customers to click past it. Start with one tactic, measure it properly, then layer the next.
Calculate your baseline AOV and segment it so you can measure changes
The formula is total revenue divided by number of orders over a set period. Say a store takes $50,000 in revenue from 500 orders in a month: AOV is $100. That’s the number everything else gets measured against.
A single store-wide figure hides what’s actually happening. Split it by new versus returning customers (returning buyers usually spend differently), by channel (paid social orders often skew smaller than email-driven ones), by device, and by product category. Without that split, a lift in one segment can mask a drop in another and you’ll call a flat result a win.
Modal order value, the single most common order size, matters more than the average here. If your average is $100 but your mode is $65, a threshold built around the average will sit miles from where most customers already are and won’t shift behaviour at all.
Product bundling: when to bundle, bundle types and how to price them
Common bundle types cover most use cases: complementary solution bundles (a skincare cleanser plus serum sold as a “routine”), quantity or volume bundles (buy three, save on the set), and buy-more-save-more tiers that reward bigger baskets without naming a hard discount. Shopify Australia points to Lvly, which expanded its bundled offerings and saw a 6% year-on-year AOV increase as a result.
The mechanic that makes bundles work is visible maths, not disguised discounting. Show the individual prices, show the bundle price, let the saving be obvious at a glance. Customers who have to calculate the saving themselves usually don’t bother, and the bundle sits there unconverted.
- Show individual SKU prices alongside the bundle price so the saving is instantly readable.
- Use dynamic or virtual bundles where your platform supports it, rather than creating a new SKU for every combination.
- Limit each bundle to products that solve one job together, not a pile of unrelated stock you’re trying to clear.
Pro Tip: Keep each bundle focused on products that solve one job together. Beyond that, the perceived value drops because the offer starts looking like a clearance bin rather than a curated set.
Avoid catalogue bloat. Every new bundle SKU is another listing to maintain, price and keep in stock, and a vendor case from BigCommerce’s Smart Bundles AI specifically flags zero catalogue bloat as a selling point of dynamic bundling tools over static ones.
Upsell and cross-sell best practice: placement, messaging and product attributes
Placement decides whether a recommendation gets seen or ignored. Product pages work for discovery, where the shopper is still deciding. Cart pages work for conversion, where intent is already set and you’re adding to a basket that’s about to close. Post-purchase screens work for low-friction add-ons, because the customer has already committed to buying and the barrier to one more click is tiny.
Baymard’s research found 68% of desktop sites are missing at least one of six attributes that recommendation cards need to actually convert:
- A high-quality product thumbnail.
- A fully visible product title (not truncated).
- A clear, visible price.
- An average rating.
- The number of ratings behind that average.
Including these details gives shoppers useful information without requiring them to click through.
For low-cost attachments, show the item with minimal copy and let the price do the convincing. For premium upsells, lead with the benefit line and the rating count. Test which framing wins for your own catalogue rather than assuming one approach fits every price band.
Spend thresholds and incentives: how to pick the right free-shipping or gift cutoff
Stripe’s guidance is specific: set free-shipping or gift thresholds 20 to 30% above your modal order value, not your average. Set the threshold close enough that shoppers can see the gap and close it with one more item.
Set the threshold against the average instead and you’ll likely price it out of reach of the customers who were closest to crossing it.
Free shipping tends to beat a percentage discount because it removes a cost the customer already resents, rather than just shaving a few dollars off a price they’d already accepted. Gift-with-purchase works well for categories where a small, relevant freebie feels like a genuine bonus rather than padding. Percentage discounts are the blunt instrument, useful for clearing stock but easiest to abuse.
Stripe’s own caveat is worth repeating: rotate the incentive and test it, because customers who learn there’s always a threshold to hit will simply wait for it, and watch conversion alongside AOV so a bigger basket doesn’t come at the cost of fewer baskets overall.
Personalisation and AI-driven recommendations: what to personalise and where
Generic “customers also bought” carousels underperform because they ignore what’s actually in front of the shopper. The signals worth using are cart context (what’s already in the basket), session behaviour (what’s been viewed in this visit), past purchase history, and which channel brought the customer in, since a paid social visitor and an email subscriber arrive with different intent.
Apply these signals on product pages, cart pages, in post-purchase emails and in post-purchase upsell flows. Baymard’s cart research found that recommendations tied closely to what’s already in the cart get meaningfully more engagement than generic suggestions sitting in the same spot, because relevance is doing the work that a bigger discount would otherwise have to do.
The mechanism is simple: a recommendation that matches context reads as help. A recommendation that doesn’t reads as noise, and gets scrolled past.
Post-purchase offers and checkout protection
A useful moment to test an upsell is right after the order confirms, when the customer has already committed to buying. A single, relevant one-click add-on here may be worth testing against offers shown earlier in the funnel.
Protecting that moment matters as much as building it. Guest checkout, saved payment methods and buy-now-pay-later options all reduce the friction that kills bigger baskets at the last step, something we cover in more depth in our checkout optimisation guide. Never force account creation or make a customer re-enter card details to accept a post-purchase offer: that single extra step is enough to lose the add-on entirely.
Measure, test and iterate: a practical test plan
Run one change at a time. Test it until you’ve got enough orders in each segment to trust the result, and track conversion rate alongside AOV every time, never AOV alone.
- Set the baseline AOV and conversion rate for the segment you’re testing.
- Launch one change: a bundle, a threshold, a recommendation placement.
- Segment results by new versus returning, channel and device before you call a winner.
- Check margin, not just revenue: a bigger basket at a deeper discount can be a net loss.
- Scale only the changes that hold up once conversion and margin are both in the picture.
A lift in AOV that costs you conversion isn’t automatically bad. If the margin on the extra basket size outweighs the drop in order count, it’s still a win, but only check that against your own numbers rather than assuming it.
Pro Tip: If you’re testing across several segments at once, a dedicated platform like VWO or Optimizely keeps the segment splits clean instead of relying on manual spreadsheet tracking.
Implementation checklist and tool categories to deploy quickly
Move in this order: measure baseline, pick one quick win, run the test, review the segmented result, then scale.
- Recommendation engines for contextual product suggestions on product and cart pages.
- Post-purchase one-click tools for the add-on moment right after checkout.
- Checkout optimisation platforms to protect speed as basket size grows.
- A/B testing and analytics tools to segment and validate every change before it’s rolled out store-wide.
Each step needs a different owner: product ops for the catalogue and bundle setup, creative for the recommendation copy and imagery, analytics for the segmented read, and someone watching the customer data signals that feed personalisation.
How we think about AOV lifts
We treat AOV as a sequencing problem, not a single tactic. Conversion optimisation, creative and campaign targeting all move the same number, and testing them in isolation misses how they interact. Happy to share the checklist template we use internally if it’d help you get started.
— Harry
Want a hand running this properly
Running bundling, threshold testing and post-purchase offers all at once, while still watching conversion and margin, takes more hours than most in-house teams have spare. We pull conversion optimisation, ad creative, integrated campaign work and email and SMS marketing under one team so none of it gets tested in isolation and dropped the moment someone goes on leave.
A consult with us can produce a plan with the quick wins flagged first: the threshold that’s sitting in the wrong spot, the cart page missing three of the six attributes a recommendation card needs, the post-purchase flow that doesn’t exist yet. What we offer covers:
- Conversion optimisation: fixing the placement and attribute gaps costing you attach rate right now.
- Ad creative production: creative built to bring in customers who already fit your higher-AOV segments.
- Integrated campaigns: strategy, creative and paid media aligned so acquisition and AOV work move together.
- Email & SMS marketing: post-purchase flows and threshold nudges that reach customers after the sale, not just before it.
If you’d rather start with a conversation than a proposal, book a consultation and we’ll walk through where your store’s quickest wins actually sit.
FAQ
How can I increase my online sales?
Focus on the levers that lift order size rather than just traffic: bundling complementary products, contextual cross-sells at checkout, and a spend threshold set just above your modal order value. Pair that with post-purchase one-click add-ons, since that moment converts at a higher rate than almost any other point in the funnel.
What is an average order value?
Average order value is total revenue divided by the number of orders over a given period, giving you the typical dollar amount spent per transaction. It’s most useful when segmented by new versus returning customers, channel and device rather than tracked as one store-wide figure.
What are the 4 C’s in sales?
Definitions of the “4 C’s” vary depending on the source and the discipline being described, covering anything from customer-focused selling principles to broader marketing frameworks. There’s no single agreed version specific to AOV or ecommerce, so treat any claimed definition with caution.
What does AOV stand for?
AOV stands for average order value, the typical amount a customer spends in a single transaction. It’s calculated as total revenue divided by total number of orders.
How much can bundling realistically increase AOV?
Results vary by category and how the bundle is priced, but Shopify Australia cites a real example where Lvly saw a 6% year-on-year AOV increase after expanding its bundled offerings. Treat that as an example, not a guaranteed outcome, since bundle design and product fit both affect the result.
Sources
- 6 List Item Attributes to Include for Cross-Sell Recommendations (68% of Desktop Sites Are Missing One or More) – Baymard
- How to Increase Average Order Value (AOV) | Stripe
