Product journey map for Shopping and Performance Max

A product journey map connects the product that attracted an ad click with the products the shopper later viewed and bought. For Shopping and Performance Max, it answers a practical question: does this product sell itself, introduce purchases of other products, or do both?

In the illustrative example below, a tee that sells little on its own opens 20 orders worth €680.00 of contribution, most of it from a coat.

How do you separate the entry product from the purchased product?

Keep the originally clicked product alongside the purchased order lines. Replacing the entry identifier when someone changes products hides the relationship you are trying to understand.

For example, a shopper clicks a tee ad, views a coat and buys the coat. The tee is the entry product; the coat is the purchased product. Recording only tee sales understates what happened after entry. Recording a tee purchase would be incorrect.

Google’s ecommerce measurement guidance distinguishes event-level and item-level metrics. Treat the event and its individual products as different levels of information when reconciling a journey.

Use stable product identifiers, with an explicit mapping where the feed and shop use different IDs. A title is a poor join key: wording can change, and variants can share similar names.

How do you count orders without losing basket detail?

Use mutually exclusive order groups when describing purchase outcomes. An order containing both the tee and coat belongs in one combined group, rather than appearing as two separate orders.

In this illustrative cohort, all 20 orders followed entry through the tee. Five contained only the tee, 12 only the coat, and three contained both. This is an order breakdown, not a conversion rate: we have not specified the number of entry visits.

Show the dataHide the data

The entry product is not always the purchased product

Purchased order contents	Orders
Tee only	5
Coat only	12
Tee and coat	3

The tee appears in eight orders, and the coat in 15, but there are still only 20 distinct orders. Summing product-presence counts gives 23 because three baskets contain both items.

When calculating a rate, state the denominator and observation window. “Orders per eligible entry visit” and “share of orders containing the coat” answer different questions. Keep unmatched and still-open journeys visible rather than treating every missing purchase as a confirmed abandonment.

What is a product journey worth in contribution?

Judge the entry product against the economics of the connected orders, while keeping each order’s contribution counted once. Revenue alone cannot show what remains after product and other variable costs.

Assume these same 20 orders produce the following amounts, excluding VAT. Contribution here is net revenue after refunds, less product costs and other variable order costs, before advertising and fixed overhead.

Illustrative data: total contribution is €680.00. If advertising for the same eligible entry cohort costs €400.00, the observed contribution after ads is €280.00, before fixed overhead. The calculation is €680.00 − €400.00.

That result describes the cohort under these assumptions. It does not prove that removing the tee’s advertising would remove €680.00 of contribution. Some shoppers may have bought anyway; incrementality testing addresses that different question.

For targets, start with the break-even ROAS calculator and one contribution-margin definition. Don’t apply one product’s margin to revenue earned by a different basket.

Break-even ROAS is 1 ÷ contribution margin before ads.

In profit terms, POAS is ROAS × contribution margin, with break-even at 1.0.

What should you do with each product’s role?

Investigate the entry and purchased products separately before changing priority. A product that closes a valuable downstream order may behave differently when it is the advertised entry point.

Product Journeys connects entry products with purchases of other products. Read that context alongside Inventory Insights and the products’ margins, rather than treating direct product sales as the whole explanation.

Product Segmentation segments products by advertising clicks and ROAS or POAS. Journey roles are additional context, not replacement segment definitions. A gateway product is not automatically a Star, and a downstream seller is not automatically a Dog.

Review one product family before expanding

Choose a product family where entry products and purchased products often differ. Reconcile the order lines, document the review window, and agree what evidence would justify a change.

Use permitted identifiers and record gaps in the observed paths. Do not join interactions merely because two shoppers bought similar products.

Keep questions about post-purchase experience and repeat purchases separate from the initial order, so later outcomes do not get counted as additional first orders.

When the evidence supports a priority change, use approved Merchant Center labels and configured campaigns to apply it. Google’s custom-label documentation explains their grouping role; a label does not create a manual per-product bid control in Performance Max.

Final thoughts

A product’s own sales are only part of its value; the orders it opens are the rest.

Start with one decision: whether an entry product deserves higher priority, lower priority or further investigation. Then review the downstream result using the same order and contribution definitions. The Performance Max guide explains how that product review fits the wider campaign structure.

For the wider customer-experience view, Hanover Research’s 2022 customer journey mapping report covers journey mapping across touchpoints and teams.