Product Bundling Around a Bestseller to Raise Order Value

A bundle raises order value when you build it from what customers already buy with a steady bestseller. Take the add-on from your order history, keep the discount below the add-on’s gross profit, and offer the bundle on the bestseller’s page and in the cart. Keep it out of Shopping ads at first: the bestseller’s ads already bring buyers, and in the feed a bundle is a new product with no ad history. Advertise it once it earns more gross profit per order than the bestseller alone.

A product bundle puts several different products on one page at one price, usually a main product with add-ons. It is one of the order-value levers in our guide to pricing on Google Shopping, next to price against the market, discounts and shipping.

Why build a bundle around a bestseller?

A bestseller already has demand, and you pay for its clicks anyway. A bundle offers those shoppers a bigger basket at the moment they decide to buy. You work with the traffic and the products you already have.

The effect shows up in average order value: revenue divided by the number of orders. ROAS, the revenue you get for each unit of ad spend, splits into three parts:

ROAS = conversion rate × average order value ÷ cost per click

A bundle works on the middle part. The click costs the same, and the share of visitors who buy can stay the same, but each order brings in more. That is how a bundle can lift ROAS without a single extra click.

Average order value moved with ROAS as closely as conversion rate did

We checked how conversion rate, average order value and cost per click moved with ROAS in GetProfit data: 1,360 store-month observations, June 2025 – June 2026. We compared each store’s monthly change with the median change across all stores in the same calendar month. That way, seasons and platform shifts that hit everyone cancel out. Rank correlation (rho) shows how closely two figures rise and fall together: the higher the value, the closer the link.

Average order value tracked ROAS as closely as conversion rate did. In the 361 months when a store’s ROAS grew relative to other stores that month, ROAS was up 47.8%, conversion rate 23.3% and average order value 19.5%. Spend stayed flat (±0.0%), and cost per click fell 2.4%.

Read this as a direction, not a forecast. It is an observation, not an experiment, and it does not show what caused each change. ROAS here is revenue-based, as in Google Ads, so it says nothing about profit. The full comparison of the two levers is in order value or conversion rate.

Which bestseller should carry the bundle?

A product that topped your sales last month is a weak base. In our study of 1.4 million products, 64.7% of winning products were one-offs: one conversion in one month and nothing after. And 71% of bestsellers stayed in the top for 3 of 13 months or less. A bundle needs a main product that will still sell next quarter.

Pick a product that:

• has sold steadily through the last year, not in one spike, so the bundle has buyers after the launch;
• holds your target ROAS in ads, so the traffic you build on pays for itself;
• is in stock and will stay in stock, because a bundle ships only with its main product;
• has natural add-ons: accessories, consumables, matching items.

The portal sorts every product into one of eight labels by how it behaved in ads over the last 12 months. In the portal’s products section, the Bestsellers label marks products that sell consistently and hold the target ROAS. That list is your shortlist for the first bundle.

How do you find the right add-ons?

Start with your order history. The question is what people already buy together with the bestseller.

1. Export the last few months of orders that contain the bestseller.
2. For every other product in those orders, count the share of bestseller orders it appears in.
3. Keep the products that complement the main product, cost less and are in stock.

Google Ads gives a second view if you send conversions with cart data. According to the Google Ads API documentation (Upload click conversions), Google Ads needs cart data to report the items purchased, their price and quantity.

The Google Ads API reference (Metrics) counts a sale as cross-sell when the product bought differs from the product in the Shopping ad clicked before the purchase. For the bestseller’s ads, cross-sell is what else people bought after clicking them. Our guide to the cart items cross-sold report shows how to read it.

Run one check before you discount anything. If many bestseller orders already include the add-on at full price, a bundle discount mostly hands money back to people who would have paid in full. Then pick a different add-on, or offer the bundle at full price.

Bundle, multipack, manufacturer’s set or checkout deal: which are you building?

People call four different things a bundle, and the feed treats each one differently.

Google’s Merchant API reference (ProductAttributes) describes both attributes for groupings the business defines itself. is_bundle covers different products sold together at a single price, and multipack gives the number of identical products in the pack. This article covers the first row.

How do you price a bundle so it earns more than the bestseller alone?

A bundle pays off when it earns more gross profit per order than the bestseller sold alone. So keep the discount below the add-ons’ gross profit. Discount more, and every bundle sold to someone who would have bought the bestseller alone earns you less.

Example store, not client data.

A tableware shop with 3,000 products spends 40,000 a month on Google Ads and gets 300 orders worth 180,000. That makes an average order value of 600 and a ROAS of 4.5. Its gross margin is 35% of revenue. A teapot at 900 is its bestseller, and two matching cups, 500 for the pair, are the most common add-on in teapot orders.

Before any discount, the bundle leaves (900 + 500) × 0.35 = 490 of gross profit. The teapot alone leaves 900 × 0.35 = 315. The break-even discount equals the cups’ gross profit: 500 × 0.35 = 175. A 20% discount on the whole bundle looks modest to the shopper and loses 105 on every order compared with the teapot alone.

The discount also goes to shoppers who would have bought the teapot and cups at full price anyway. At 1,300, each buyer who switches from the teapot alone to the bundle adds 75, and each of those full-price buyers costs 100. The bundle pays off as long as switchers outnumber full-price buyers by more than four to three.

If 60 of the month’s 300 buyers take the bundle at 1,300 instead of the teapot alone, revenue grows by 60 × 400 = 24,000, to 204,000. Average order value rises from 600 to 680 and ROAS from 4.5 to 5.1, on the same 40,000 of spend and the same clicks. Gross profit grows by 60 × 75 = 4,500.

Should the bundle go into Shopping ads right away?

Start on your site and add the bundle to ads later, for three reasons.

1. The traffic is already there. Shoppers who click the bestseller’s ad land on its page. A bundle offer on that page and in the cart works on those clicks. If your conversion tag sends the real order value, the bigger order reaches Google Ads whether or not the bundle has its own ad.
2. A new product starts from zero. In the feed, the bundle is a separate product with its own ID and an empty ad history. Our study of 1.4 million products tracked 1,196,357 products that began the 13 months with zero spend: 74.1% of them never got a single impression. Those that went on to sell took a median of 4 months to their first sale. A bundle pushed straight into ads can stall the same way.
3. You learn whether it sells before you pay for clicks on it. A month on the site shows how many bestseller buyers take the bundle and what it does to gross profit per order.

If you send cart data, add the bundle to the feed anyway. The Google Ads API reference (CartData.Item) requires each sold item’s ID to equal its Merchant Center product identifier, so a bundle missing from the feed has nothing to match.

To keep the bundle out of ads until the test is over, exclude it from your campaigns. In Performance Max, listing groups decide which products each asset group includes or excludes. Google’s documentation (Listing Groups for Retail) recommends targeting groups of products, for example by custom label or brand, rather than single item IDs. So a custom label in the feed, such as “bundle-test”, is the simplest switch.

When the bundle is ready for ads

Move the bundle into ads once it sells on the site and earns more gross profit per order than the bestseller alone. Then get the feed entry right:

• is_bundle = yes. It tells Google the product is a bundle you put together yourself: different products sold at one price.
• Title, image and description for the whole bundle, not only the main product, so the listing shows what the shopper gets.
• The right identifiers. Before launch, check in Merchant Center Help which GTIN, MPN and brand a bundle you put together should carry. Check there, too, whether your target countries require is_bundle.
• Identical products make a multipack, not a bundle. They take the multipack attribute instead.

For the feed side, including which identifiers each type carries and where the attributes are required, see bundles and multipacks in the feed.

Once the bundle runs in ads, treat it as a new product. The portal labels products with no ad history as Newcomers and advises giving them first impressions in a test campaign. Keep the bundle in its own product group, so its numbers stay separate from the bestseller’s.

How to launch your first bestseller bundle

1. Pick a steady bestseller: it sold through the last year, holds your target ROAS and is in stock. In the portal, start from the Bestsellers list.
2. Pull the last few months of orders that contain it and rank the other products by how often they appear with it. This shows what buyers already want with it.
3. Choose one or two add-ons that complement it, cost less and are in stock. One clear offer is easier to accept than a menu.
4. Check how many bestseller orders already include those add-ons at full price. Those buyers get the discount on a basket they would have bought anyway.
5. Set the price so the discount stays below the add-ons’ gross profit. Otherwise each bundle earns less than the bestseller alone.
6. Offer the bundle on the bestseller’s page and in the cart. Leave your campaigns as they are, so the result shows the effect of the bundle alone.
7. After a full month, compare bestseller orders before and after: the share taking the bundle, average order value, gross profit per order and the bestseller’s own conversion rate. A bundle that lowers the bestseller’s conversion rate can cost more than it adds.
8. If the bundle wins, add it to the feed with is_bundle, the identifiers Merchant Center Help asks for, and a title and image for the whole bundle. Then test it in ads as a new product.

See which of your products sell and which only spend. Every product gets a label based on how it behaves in ads. The portal changes nothing without your consent.

Check your catalogue →

Sources

• Upload click conversions (Google Ads API) — cart data is required to report items purchased, price and quantity in Google Ads. Last updated 30 September 2026, checked 2 October 2026.
• Metrics (Google Ads API reference) — definitions of cross-sell and lead sales and of average order value with cart data. Last updated 30 September 2026, checked 2 October 2026.
• CartData.Item (Google Ads API reference) — a sold item’s ID must equal the Merchant Center product identifier. Last updated 22 July 2026, checked 2 October 2026.
• ProductAttributes (Merchant API reference) — is_bundle for business-defined bundles of different products at one price; multipack for the number of identical products. Last updated 23 September 2026, checked 2 October 2026.
• Listing Groups for Retail — listing groups include or exclude products per asset group; groups of products, for example by custom label or brand, preferred over item IDs. Last updated 30 September 2026, checked 2 October 2026.
• GetProfit data: 1,360 store-month observations, June 2025 – June 2026 — ROAS linked to average order value and conversion rate; changes in months of ROAS growth.
• GetProfit study: 1,404,808 products in 130+ stores over 13 months — bestseller lifespan; products that started with zero spend.