If you run an operating print-on-demand store, you don't need convincing that bundles can lift average order value. You need to know which bundle, at which price, without handing away margin you can't get back.
That's a data problem, not a creativity problem. And the data you need already lives in your store — it's just scattered across orders, supplier invoices and ad reports that rarely sit side by side.
What bundling analytics actually answers
Cross-sell and recommendation engines are a huge revenue lever — roughly a third of Amazon's revenue is attributed to its recommendation and "frequently bought together" engine, a McKinsey figure cited across ecommerce. Bundling is you doing that work deliberately instead of hoping an algorithm does it.
Good bundling analytics answers four questions in order:
- Affinity — which two or three products appear in the same order more often than chance?
- Margin — after the bundle discount, what is the contribution margin per bundle?
- Incrementality — is the bundle creating a bigger basket, or discounting a basket you'd have earned anyway?
- Inventory and fit — does the bundle move a slow SKU without starving a fast one?
Skip any of the four and you get a bundle that looks popular and feels like a win while trimming your profit per order. The rest of this guide walks each one with real numbers. For the wider picture of how these metrics connect, our guide to ecommerce business intelligence maps the full stack.
Step one: affinity — find the pairs, don't guess them
Affinity analysis (sometimes called market basket analysis) looks at which items land in the same cart together. The honest version compares the joint rate to what you'd expect from each product's standalone popularity.
Say your best tee appears in 40% of orders and a matching mug in 10%. If the two were unrelated, you'd expect them together in about 4% of orders (0.40 × 0.10). If they actually show up together in 12%, that's three times the chance baseline — a real affinity, not just two popular products colliding.
That lift-over-baseline is the signal. A pair that's "often bought together" only because both are bestsellers isn't a bundle opportunity; it's two things everyone already buys. You want pairs where buying one genuinely pulls the other.
Step two: the margin math every bundling guide skips
Here's where most articles wave their hands. Let's not.
Say you run a POD apparel store doing around 900 orders a month at a $38 AOV. You want to bundle a graphic tee and a matching hoodie. Separately they look like this:
- Tee: price $28, supplier cost $12
- Hoodie: price $45, supplier cost $24
Bought separately that's $73 in revenue against $36 in product cost — $37 of gross profit. Because Printify, Printful and Gelato give you the blank-plus-print cost before you ever list the item, you know that $36 on day one, not after the fact.
Now you bundle the pair at $65 — an $8 discount. Product cost is still $36, so gross profit drops to $29. Then subtract the variable costs of actually shipping it: one package at $5, payment processing at 4% ($2.60), and pick/pack at $1.40. That leaves a contribution margin of $20 per bundle, or about 31% of the $65 price.
The $8 you discounted came straight out of margin — there's no supplier rebate for bundling. So the bundle only wins if it changes behavior, which is the next question.
Step three: incrementality — the trap that eats profit
A bundle that a customer would have assembled anyway, at full price, is just a coupon you mailed yourself. The analytics question is whether the bundle moves someone up.
Walk it through. A customer who buys only the tee at $28 earns you roughly $8.50 in contribution margin (price $28 − cost $12 − shipping $5 − processing $1.12 − pick/pack $1.40 ≈ $8.48). If the bundle converts that single-tee buyer into a $65 bundle buyer, your contribution jumps from about $8.50 to $20 — an extra ~$11.50 on that order.
But if the customer was already going to buy both items separately, the bundle costs you. You'd have earned contribution on $73 of full-price revenue; now you earn it on $65. The same bundle is a win or a loss depending entirely on who's buying it.
That's why you segment: compare bundle buyers against your old "bought both separately" rate. If the share buying both items barely rose but your blended margin fell, the bundle is cannibalizing. If more customers now carry two items where they used to carry one, it's incremental — and worth scaling.
Step four: inventory fit and the denominator traps
Bundling is also an inventory tool. Daasity's product bundling analysis lays out the classic plays: pair a high-margin slow mover with a fast seller, or combine items customers already buy in sequence into one larger first order.
Watch your denominators, because this is where bundle "wins" get faked:
- Attach rate — bundle units ÷ eligible orders, not total orders. If only half your traffic sees the bundle, measuring against all orders understates it.
- Bundle AOV vs blended AOV — a bundle lifts the orders that contain it; that's not the same as lifting store-wide AOV if the bundle is rare.
- Discount leakage — revenue booked before refunds and returns overstates every bundle. Net returns out before you call one profitable.
Cart abandonment is the other quiet denominator. The Baymard Institute puts the documented average at 70.22% across 50 studies — so a bundle that raises cart value but also raises sticker shock can lower completed orders. Track the bundle to the completed, net-of-returns order, never the add-to-cart. Our ecommerce performance analytics guide goes deeper on tying these funnel metrics back to profit.
Does a bundle actually lift AOV?
Often, yes — but the published numbers are marketing claims, not guarantees. One widely shared case study reports bundles driving a 55% AOV increase in a best case; your own store will land wherever your affinity and pricing land it.
Treat any outside AOV-lift figure as a reason to test, not a number to expect. The only lift that matters is the one your own orders show after the discount — which is exactly the measurement the four steps above produce.
Why this is hard with a dashboard, and what Victor does
The reason bundling analytics stalls isn't the formulas — it's that the inputs live in different places. Your order and discount data sit in Shopify, your true product cost sits with Printify, Printful or Gelato, and your acquisition cost sits in Meta Ads and Google Ads. A dashboard shows you each one; it won't do the join that tells you the bundle's real profit.
PodVector AI's Victor is an AI employee, not a dashboard. Victor sees every order, every ad dollar and every supplier cost together, so he can tell you the true per-order profit on a bundle and whether its buyers are incremental or cannibalizing full-price sales.
Because Victor works across Shopify, Meta Ads, Google Ads and Klaviyo, he can also act on what the data says — with your approval. He can set up the bundle's price and discount in Shopify, build a Klaviyo flow to promote it, and adjust the ad budgets behind it; every write action waits for you to approve it before anything executes. He delivers the profit read as a report to your Google Drive so you're not rebuilding the math each week.
See your true per-bundle profit with Victor
If you're comparing tools first, our rundown of the top ecommerce analytics platforms for 2025 and the Shopify developer dashboard covers where each one stops.
FAQs
What ecommerce data do I need to build a bundling strategy?
Four feeds: order history (to find affinity and attach rate), product cost (your supplier's blank-plus-print price), variable fulfillment costs (shipping, payment fees, pick/pack), and ad spend by product or campaign. Affinity alone tells you what pairs; the cost and ad data tell you whether the pair is worth bundling.
How do I know which products to bundle together?
Compare how often two items appear in the same order against the rate you'd expect from their individual popularity. A pair bought together far more than chance predicts has real affinity. Then confirm the combined bundle still clears a healthy contribution margin after the discount — affinity without margin is a trap.
How much should I discount a bundle?
Start from the margin, not a round percentage. Calculate the contribution margin of the bundle at several price points and pick the largest discount that still clears your target margin per order. In the worked example above, an $8 discount left $20 of contribution — fine if buyers are incremental, costly if they'd have bought both anyway.
Does bundling actually increase profit or just revenue?
It depends on incrementality. A bundle that turns single-item buyers into multi-item buyers raises both. A bundle that mostly discounts customers who'd have bought everything at full price raises revenue while lowering profit per order. You can only tell the two apart by watching contribution margin and the share of orders carrying multiple items — not revenue alone.
Can PodVector AI help me price and launch a bundle?
Yes. Victor computes the true per-order profit on a proposed bundle using your live Shopify, supplier and ad data, then — with your approval — can set the bundle price and discount in Shopify, build the Klaviyo email flow, and adjust the ad spend behind it. Every change is approval-gated, so nothing goes live until you say so.
What's the difference between bundle AOV and store AOV?
Bundle AOV is the average value of orders that contain the bundle; store AOV is the average across all orders. A bundle can have a high bundle AOV and still barely move store AOV if few orders include it. Judge a bundle on its contribution margin and attach rate, then watch store-wide AOV to see whether it's meaningfully changing the business. For the B2B angle on these metrics, see our B2B ecommerce analytics guide.