What this playbook covers and when to run it
The playbook covers the operating life of product advertising in Google Ads: setting the structure from the catalogue and its margins, keeping the Merchant Center feed clean, moving products through their lifecycle, refreshing assets and signals, managing bidding and budgets, reconciling platform results with the store's own orders and testing what Performance MaxA Google Ads campaign type that runs one goal across all Google inventory from shared assets. adds. It does not teach any of those methods. The Shopping and Performance Max hub explains why the feed is the campaign and which campaign type to run, the guide to shopping and catalogue ads gives the cross-platform view, and each phase below hands off to the guide, checklist or experiment that owns the method.
It is the product-campaign counterpart of the paid search account management playbook, which runs search campaigns: search terms, negatives, ads and landing pages. The Google Ads account hub sets a management routine for the whole account; this playbook is the part of it that concerns products. Asset production runs through the creative production playbook, and sales periods, with their stock, promotion and service decisions, through the peak season trading playbook.
Run it when any of these is true: the same disapprovals return every week; nobody can say which products the campaigns are pushing or why; targets are the same for every product whatever its margin; clearance and new lines share one campaign; asset groups have not changed since launch; or the platform's revenue and the order system's revenue are never compared. Like every playbook in the playbooks library, it is a starting shape: a small shop may run it in a few hours a week, while a large catalogue will spread it across ecommerce operations, merchandising, creative and the paid media team.
Roles and responsibilities
- Shopping and Performance Max manager. Owns the campaigns: structure, listing groups, asset group set-up, bidding strategies, targets and budgets. Reads alerts and pacing daily and decides what the campaigns do with each product group.
- Feed owner. Owns the product data in the store platform and Merchant Center: attributes, titles, identifiers, variants, custom labelsFree-text product feed attributes a store fills in to group its products for campaign structure, bidding and reporting, such as margin tier or season., diagnostics and the sync of price and availability. Fixes problems at the source wherever possible.
- Merchandiser. Owns the commercial view: margins, stock depth, launches, seasonal ranges and clearance. Decides which products matter and tells the feed owner and the campaign manager when that changes.
- Creative lead. Owns the assets in each Asset GroupThe Performance Max unit holding the text, images, videos and signals for one theme, product set or audience.: text, images, logos and video, briefed and refreshed through the creative production routine.
- Analyst. Owns ecommerce tracking, the monthly reconciliation with orders and the design and reading of incrementality tests.
In a small business one person may run the campaigns and the feed, and the merchandiser may be the owner. The split still matters, because the commonest failure is a campaign manager making merchandising decisions by default, or a feed owner changing data the campaigns depend on without telling anyone.
Cadence, inputs and outputs
The rhythm below is a suggested starting point. Adjust it to the size of the catalogue, how fast products and prices change and how much of the budget runs through product campaigns; the feed guide itself suggests running its diagnostics more often during sales periods.
| Cadence | Owner | Inputs | Outputs |
|---|---|---|---|
| Daily | Campaign manager | Account and Merchant Center alerts, spend against budget | Urgent account or feed issues raised to the feed owner, budgets corrected, notes logged |
| Weekly | Feed owner, campaign manager, merchandiser | Merchant Center diagnostics, product and stock changes, search terms and insights | Top disapproval reasons fixed at the source, labels updated, product decisions recorded |
| Monthly | Campaign manager, analyst, merchandiser | Results by margin band and lifecycle stage, the order system's figures, the margin data | Structure, targets and budgets adjusted from margin; a reconciliation ratio tracked |
| Quarterly | Creative lead, campaign manager, analyst | Asset reports, range changes, the test backlog | An asset refresh plan and one incrementality question designed or read |
The daily read catches what stops products serving; the weekly cycle keeps product data and product decisions in step; the monthly review sets targets on numbers the finance team recognises; and the quarterly work keeps assets fresh and checks that the campaigns add what they claim. Results reach the business through the team's existing monthly and quarterly packs.
The playbook in phases
Set the structure from the catalogue and its margins
Split campaigns and listing groups where margins, priorities or stock differ, not where the catalogue's menu happens to divide. The Standard Shopping structure guide covers product groups, campaign priority and negative keywords, and when Standard Shopping is the right container for products that need query control. The Performance Max structure guide covers how many campaigns to run, asset and listing groups, brand and placement exclusions, and the precedence rules Google documents for running beside Search and Shopping. The performance economics guide sets targets from unit economics; the break-even calculator turns gross margin into the return below which sales lose money. New campaigns pass the Performance Max launch checklist before they are enabled.
Keep the feed clean
The feed owner runs the weekly diagnostics routine in the product feeds guide: account, feed and item issues in that order, disapprovals counted by reason, price and availability mismatches, limited items and new items checked, with the item list ordered by revenue at risk. The guide's rule is to fix problems at the source where the source is the store platform, and by rule or supplemental feed only where it cannot change quickly; a disapproval that recurs after a fix was fixed downstream. Attributes follow the product data specification in Merchant Center Help, as the guide describes. The product variants tutorial gets colour, size and other variants right, and the Merchant Center feed set-up checklist is the gate when a feed or a store platform changes. Before the feed owner rewrites titles across a whole category, the shopping feed title test is a design for retitling a random half first and reading both halves against their own pre-period.
Move products through their lifecycle
New lines, best-sellers, long-tail products, seasonal ranges and clearance need different treatment. The product lifecycle guide labels each product's stage in one custom label slot, populated from data and refreshed on a schedule, and gives each stage its campaign treatment, including what to do with products that go out of stock or are discontinued. When stock has to be cleared, the guide to clearing excess stock keeps clearance in its own group with a target from the clearance margin, paid support only where stock justifies it, and feed prices matching the clearance pages. The merchandiser supplies the stage decisions; the feed owner writes them into the feed.
Refresh assets and signals
Asset groups tire, and ranges change under them. The tutorial on building an asset group takes a group from a written brief through text, images, logos, video, audience signals and search themes to its first asset report, and is the method for every rebuild. The creative lead plans refreshes quarterly, as a starting rhythm, from the asset reports and the merchandiser's calendar, and briefs them through the creative production playbook, so product campaigns get assets on the same footing as every other channel.
Manage bidding and budgets
The Smart Bidding strategies guide explains what each strategy optimises, how to set target return or cost from the business's numbers and how to change targets without resetting learning. The budgets and pacing guide explains how average daily budgets are spent, what limited by budget means and how to pace a month. Sales, launches and tracking faults are unusual days that mislead automated bidding; the guide to sales and tracking outages covers seasonality adjustments, data exclusions and the changes to hold back. The ROAS calculator shows what a campaign returns against its break-even point. Targets move monthly, from margin, and never in reaction to a single day.
Reconcile with orders
Once a month, as a starting rhythm, compare what the platform reports with what the store recorded. The tutorial on reconciling Google Ads with sales records explains the gap line by line, separates structural differences from faults and produces a ratio between the platform figure and the business figure to track each month and set ROASRevenue generated per unit of ad spend. targets against. The GA4 ecommerce tracking tutorial keeps the analytics side of the same purchases accurate, so a fault shows up as a moving ratio rather than a surprise at year end.
Test what Performance Max adds
A campaign's own report cannot say whether it grew sales or claimed sales other campaigns would have won. The Performance Max uplift test compares treated and control regions on total account conversion value, with brand conversions, new-customer share and blended cost per conversion as guardrails. The incrementality testing guide covers designs, sizing and reading lift. Ask one question at a time, design it before the result is wanted and decide in advance what each result will change in the budget.
Who decides about a product: merchandiser, feed owner or campaign manager
Most product campaign arguments are about who decided that a product should be pushed, held back or changed. Agree the split once and record each decision where all three can see it.
| Decision | Who decides | Who acts | Evidence |
|---|---|---|---|
| Push a product or range harder | Merchandiser, with the campaign manager on cost | Feed owner sets the label; campaign manager adjusts the group | Margin, stock depth, the commercial calendar |
| Exclude or pause a product | Merchandiser for commercial reasons; campaign manager for poor returns against margin | Campaign manager excludes it in the listing group | Results against break-even, stock, policy status |
| Change titles, attributes or variants | Feed owner, with the campaign manager's search insight | Feed owner, at the source | Diagnostics, search terms, the product data specification as the feed guide describes it |
| Change a target or budget | Campaign manager, within the limits agreed with the budget holder | Campaign manager | Margin by group, the reconciliation ratio, pacing |
| Rebuild an asset group | Creative lead and campaign manager together | Creative lead briefs; campaign manager builds | Asset reports, range changes |
| Run an incrementality test | Analyst proposes; budget holder approves | Analyst and campaign manager | A written question, design and decision rule |
Two rules hold the table together. The campaign manager does not make merchandising decisions by default: a product excluded for poor returns is reported to the merchandiser, who may have a reason to keep it. And the feed owner does not change a label, title or attribute the campaigns depend on without telling the campaign manager, because a label is a campaign structure even though it lives in the feed. The academy checkpoint on running, auditing and testing a Shopping programme tests who decides about a product, fixing the feed at the source and refreshing labels and asset groups, beside the audit's gates and the uplift test's guardrails.
An outdoor lighting retailer runs its product campaigns
Clevrund Lighting, an invented outdoor lighting retailer with about 900 products, starts the playbook after a year of one Performance Max campaign at a single target for the whole catalogue. The merchandiser groups products into three margin bands, and the campaign manager splits listing groups to match, with targets set from each band's break-even return. The weekly diagnostics show the same 60 solar lights disapproved for a price mismatch every week; the feed owner traces it to the store platform's sale pricing and fixes it there instead of in a feed rule, and the disapproval stops recurring. Clearance lines get a lifecycle label and their own group, so they stop competing with the spring range. The monthly reconciliation settles at a platform-to-orders revenue ratio of 1.18, which becomes the basis for targets. Before moving a further £4,000 a month into Performance Max, the analyst runs one regional uplift test over eight weeks; it shows a smaller gain than the campaign's own report, and the budget moves in two steps.
Measuring the playbook
The playbook works when product decisions are made on purpose and the platform's numbers can be trusted. Operating signals show whether the rhythm holds: whether daily alerts are read, whether the weekly diagnostics run and the top reasons are fixed at the source, whether lifecycle labels are refreshed on schedule, whether the monthly reconciliation happens before targets change, and whether each quarter ends with an asset refresh plan and a test designed or read.
Outcome signals show whether the campaigns are getting better run: how many disapproval reasons recur after a fix, how much spend runs on products below their break-even return, how stable the reconciliation ratio stays from month to month, how many asset groups have been refreshed since the range last changed, and how many budget moves rest on a test rather than on the campaign's own report. Report them in the team's existing monthly and quarterly packs, beside the sales results they support. The academy capstone on an ecommerce trading operations plan for a party supplies shop practises writing this rhythm into a year's plan, with its audit fixes and uplift test, beside a peak run as one programme and a lifecycle and retention programme.
Common mistakes
- Fixing the feed downstream. A feed rule that patches a store problem has to be maintained forever. Fix at the source where the store platform allows it.
- One target for the whole catalogue. Products with different margins need different targets. Split by margin band or lifecycle stage and set each target from its economics.
- Labels set once. Margins, stock and seasons change every week. Refresh labels from data on a schedule.
- Campaign managers merchandising by default. Excluding a product is a commercial decision too. Tell the merchandiser, who may have a reason to keep it.
- Asset groups left from launch. A group built for last season's range keeps serving. Plan refreshes and brief them like any other creative.
- Trusting the campaign's own credit. Performance Max reports what it touched, not what it caused. Reconcile with orders every month and test before moving large budgets.
- Changing targets in reaction to a day. Unusual days mislead automated bidding. Change targets on the monthly review, from margin, and treat sales and outages with the tools built for them.
