Google Shopping ads: setup and checks in 2026
Get your product range ready for Google Shopping ads. Set campaign controls, check purchase tracking and diagnose ads that are not showing.
- Published
- Last updated
Google Shopping ads are paid product ads built from your Merchant Center product data. To launch a controlled campaign, connect eligible products, choose the inventory deliberately and confirm that purchases carry the right order values. An approved product alone doesn’t prove that its campaign can show ads.
The worked example below uses a fictional Australian homewares retailer and a Standard Shopping campaign. Its aim is to advertise two online lamp variants while keeping store-only stock and unwanted products outside the campaign.
Match the campaign route to your product range
Choose Standard Shopping when you need product-group selection and campaign controls for a product-led advertising campaign. Performance Max combines product selection with automated bidding and delivery across Google’s advertising channels.
Standard Shopping uses product groups inside ad groups. With manual cost-per-click (CPC) bidding, you can set bids on the final, unsubdivided groups. Performance Max uses listing groups inside asset groups, with bids set automatically against the campaign objective. Its product-level reports cover a different scope from campaign-wide reporting. Google explains these distinctions in its Performance Max listing-group guidance.
Shopping matching uses your product data, including titles and product attributes. You don’t build a positive keyword list and write a separate text ad for every variant. Search-term evidence can explain which searches brought visits, but it doesn’t replace catalogue eligibility or product inclusion.
Paid campaigns and free listings have separate controls. Read the Google Shopping guide when deciding whether you need paid exposure alongside free listings.
Confirm the products are eligible to advertise
Check the account connection and each intended item before creating the campaign. You need access to the retailer’s Google Ads account, its linked merchant account and the storefront’s purchase-measurement setup.
Keep these checks separate:
| Check | Expected state | Who corrects a failure |
|---|---|---|
| Account connection | Correct Merchant Center ID linked to the intended Google Ads account | Account administrator |
| Account status | No suspension or account restriction blocking Shopping ads | Account owner, following the displayed policy issue |
| Item status | Intended item eligible for the Shopping ads destination in the selected market | Product-data owner |
| Availability | Submitted availability agrees with saleable stock and the website | Catalogue or stock owner |
| Campaign inclusion | Item passes inventory filters and belongs to an included group | Campaign manager |
The Google Merchant Center guide covers account setup and access. For submission problems, use the Google product feed guide.
For fictional item LAMP-SAND, open its submitted destination. The sand variant must be selected, its displayed $120 price must agree with checkout and delivery must be available to the intended Australian locations. Repeat for LAMP-GREEN. Check the submitted currency field is AUD.
A wrong submitted price needs a product-data correction. A correct feed that opens the wrong colour needs a website or destination correction. A valid online item excluded from its campaign needs a campaign correction. Record the failure against the item ID so the right person receives it.
Create the campaign and set deliberate controls
Create a Standard Shopping campaign with an explicit purchase goal and online inventory selection. Google’s Shopping campaign instructions describe the existing-account workflow. Onboarding screens can differ.
- In Google Ads, choose New campaign. Select Sales and review the purchase goal offered for the campaign.
- Choose Shopping, then continue with Standard Shopping. Select the correct linked merchant account. Google says this merchant selection cannot be changed after creation.
- Select the feed label for the intended inventory. A feed label selects product data. Location targeting separately decides the geographic reach.
- Name the campaign Online lamps. Choose a bidding approach deliberately. This fictional example uses Manual CPC so the manager can inspect product-group bids. A click-maximising strategy pursues clicks. Value-based bidding needs reliable purchase values.
- Enter the retailer-approved average daily budget. Record the permitted spend and who can authorise changes before enabling delivery.
- Select the Australian locations the store serves. Review location options and search-partner participation against the delivery area and buying brief.
- Inspect the Inventory filter and restrict the campaign to Online. Keep store-only stock outside this online-delivery example.
- Create the Lamps ad group. Set its starting bid for the manual strategy, then replace broad product inclusion with the group rules below.
- Review the summary, purchase goal and campaign dates. Save the original settings and keep the campaign paused until product selection and measurement checks are complete.
Local inventory needs an explicit decision. Google’s local-inventory default announcement makes the inventory filter the control for separating online and local Shopping inventory. Don’t rely on an old Local products toggle to exclude store-only offers. A local campaign also needs the relevant local-inventory setup and accurate pickup or store promises.
An average daily budget isn’t a strict daily charge ceiling. For most campaigns, Google’s spending limits allow daily billing up to twice that budget. The usual full-month limit is 30.4 times an unchanged average daily budget. Budget changes and partial-month operation affect the calculation.
For a fictional unchanged $20 average daily budget, those limits are $40 for a day and $608 for a full month. These are charging limits, not a forecast of spend, purchases or profit. Confirm the campaign’s budget type before applying that example.
Organise product groups around catalogue decisions
Use product groups to include the intended lamp variants and exclude the remaining inventory. The fictional catalogue has these records:
| Item ID | Product type | Custom label 0 | Availability | Campaign decision |
|---|---|---|---|---|
| LAMP-SAND | Lighting | launch | In stock | Include |
| LAMP-GREEN | Lighting | launch | In stock | Include |
| LAMP-BLACK | Lighting | hold | In stock | Exclude |
| LAMP-WHITE | Lighting | launch | Out of stock | Include in rule, currently unable to serve |
| RUG-NATURAL | Rugs | launch | In stock | Exclude |
| LAMP-LOCAL | Lighting | launch | Store-only | Exclude through Online filter |
In the Lamps ad group, open Product groups and subdivide All products by Product type. Select Lighting. Subdivide Lighting by Custom label 0, then select launch. Within launch, subdivide by Item ID and include LAMP-SAND, LAMP-GREEN and LAMP-WHITE.
Exclude “Everything else” at every relevant branch. That catches Rugs, the hold label and additional lamp IDs outside this selected set. Before enabling, inspect the item-level result: only the two in-stock online variants are currently eligible to advertise. LAMP-WHITE remains deliberately included but unavailable until stock returns.
Google documents the supported attributes and subdivision procedure in its product-group guide. For manual bidding, review bids on the final groups after subdivision. A parent with subdivisions doesn’t take its own bid.
The catalogue owner maintains the product catalogue identities. The campaign manager checks new IDs and changed attributes against the group rules. A replacement ID won’t inherit an old ID’s explicit inclusion. A missing launch label falls into remaining inventory, which this example excludes.
Google says new or edited custom labels can take 24 to 48 hours to appear in Ads. Confirm the processed value before changing the campaign to compensate for an apparent omission.
Check purchases, values and duplicate conversions
Select the completed purchase action used for campaign decisions. Keep product views and basket additions as observational actions when purchases are the commercial goal. Review both the action’s primary or secondary setting and the goals selected by the campaign, following Google’s conversion-goal guidance.
This fictional measurement route uses a Google Ads website purchase action and the Google tag. Order HOME-1042 contains one LAMP-SAND at $120 plus $10 delivery. Assume the retailer’s agreed reporting definition includes delivery and tax, so the expected purchase value is $130.
The developer checks the completed order against the purchase event:
| Field or event | Expected result |
|---|---|
| Conversion destination | The intended Google Ads purchase action |
| Transaction identifier | HOME-1042, unique to this order |
| Value | 130, matching the agreed order-value definition |
| Currency | AUD |
| Confirmation-page reload | Same order identifier, with no extra counted purchase for the same action |
| A separate completed order | A different identifier and its own actual value |
Configure transaction-specific conversion values so each purchase sends its actual value. Sending the same default value for every basket would distort the campaign’s revenue evidence.
Google’s transaction-ID instructions explain duplicate handling for the same conversion action and transaction ID. Reusing HOME-1042 for another order can suppress a legitimate purchase. An identifier also doesn’t automatically merge separate conversion actions: avoid counting a tag-based purchase and an imported purchase as two primary sales.
In an authorised test storefront, complete an order and inspect its event payload against the saved order. Reload confirmation and inspect the repeat event. The expected result is one purchase with the saved order’s identifier and value.
A valid purchase event proves a measurement step. Advertising attribution depends on an eligible ad interaction and the action’s attribution settings. A direct test order needn’t appear as an ad-attributed sale. Allow for purchase delay and reporting processing before comparing Ads with order records. Google’s reporting-discrepancy guidance also distinguishes interaction-based reporting from conversion-time reporting.
Diagnose Google Shopping ads that are not showing
When ads have no impressions, inspect account status and campaign eligibility before raising bids. Use the item ID to follow the same product through Merchant Center and Google Ads.
| Order | Inspect | Expected state | Corrective action |
|---|---|---|---|
| 1 | Ads account alerts and Billing | Account active, billing able to fund ads | Account owner resolves the specific restriction |
| 2 | Campaign status, dates and ad schedule | Enabled within its active period and schedule | Campaign manager corrects dates or an unintended pause |
| 3 | Ad group status | Intended group enabled | Manager unpauses the correct group after readiness checks |
| 4 | Merchant Center account and item issue details | Correct destination eligible, item in stock | Product owner corrects the displayed issue |
| 5 | Locations and campaign inventory filter | Intended market and Online channel included | Manager corrects an unintended restriction |
| 6 | Product groups and item report | Exact item in an included final group | Manager corrects the group rule or replacement ID |
| 7 | Strategy status, bids and budget | Controls permit auction participation within spending authority | Manager diagnoses the constraint before proposing a change |
For LAMP-GREEN, an eligible item with zero impressions could still be outside the selected inventory. If its launch label is missing, restore the correct catalogue value and confirm processing. Raising the bid won’t repair that missing membership.
Google’s campaign-not-running guide covers account restrictions, narrow targeting and bidding limitations. Newly edited products also need review and synchronisation time. Check the reporting window before treating zero impressions as a continuing outage.
Confirm recovery in stages. First inspect the corrected item and included group. Then review subsequent impressions and clicks for that item over an appropriate reporting window.
Eligibility permits auction participation. It doesn’t guarantee every item an impression. If delivery stays absent, hand over the unresolved constraint with the original settings.
Improve the campaign and hand over specific problems
Optimise the campaign by diagnosing a product-level problem, making one justified change and comparing consistent reporting periods. Separate a measurement fault from an offer that loses money.
| Product-level signal | Next investigation | Decision it supports |
|---|---|---|
| Eligible included item, few impressions | Demand, targeting, auction constraints and strategy status | Whether a bid or targeting change has evidence |
| Clicks with irrelevant search terms | Search-term relevance and submitted product meaning | Whether exclusions or a separate product-data correction are needed |
| Clicks landing on the wrong variant | Actual destination selection and checkout | Website repair before more paid visits |
| Orders in the store, purchase value missing in Ads | Purchase action, tag payload and attribution basis | Measurement correction before judging return |
| Reliable conversion value below commercial needs | Margin after product, delivery, returns and ad costs | Offer, inventory or spend decision |
For example, suppose LAMP-SAND receives clicks but opens the black variant. Record the destination error and pause the affected item if the retailer authorises it. Repair and recheck the sand destination. Keep the original bid and budget, then compare performance after the correction using the same purchase action and value definition. Raising spend while the landing page remains wrong would obscure the diagnosis.
Give campaign manager Alex a daily alert review for account restrictions, item eligibility and unexpected spend. Use a weekly item-level review for delivery, purchase-record quality and destination mismatches. Record these as the fictional retailer’s operating choices, not universal intervals. Compare performance only after allowing the account’s normal conversion delay.
An alert opens an investigation. Alex confirms the issue in the account or order record, assigns its correction and closes it only after checking the corrected state. Keep the previous setting, reason, expected effect and next review point in the change log. A return-on-ad-spend target must come from the retailer’s economics, not a universal benchmark.
A specialist handover needs the affected IDs, merchant and campaign identifiers, eligibility messages, group membership and original controls. Add measurement evidence, the actual destination mismatch and the decision required. Keep customer details out of shared examples. The retailer retains account access and authority to approve spending.
Compare specialist management services
Compare Google Shopping management services by responsibilities and evidence, then by fees. A service can manage eligibility checks and campaign controls while coordinating measurement fixes, monitoring and reporting. Agree separately who owns catalogue facts and fulfilment promises.
An agency can provide several specialists and cover for absences, with an explicit day-to-day contact. An independent specialist can suit a narrower remit with direct access to the operator. An internal team keeps daily decisions close to stock and margin changes, but needs time and the relevant skills. These are staffing choices, not guarantees of campaign results.
Ask each proposal to specify:
- Retailer-owned Ads and Merchant Center accounts, permitted access and the exit handover.
- Who approves budgets, bids and new campaigns, including escalation when spending approaches an agreed threshold.
- Fee basis, contract terms and included work, with advertising spend shown separately.
- Deliverables for eligibility, purchase measurement and item-level reporting.
- Reporting definitions for purchase value, attribution, returns and management fees.
- Named owners for feed corrections, website repairs and fulfilment changes.
- Monitoring frequency, incident response and the evidence supplied when closing a problem.
- Exportable settings, change history and removal of access when the engagement ends.
Give every provider the same LAMP-GREEN inclusion problem and HOME-1042 measurement record. Compare the proposed diagnosis and correction ownership. A promise of guaranteed returns doesn’t establish that either problem will be resolved.
Compare Google Shopping with Facebook Ads
Choose a product-led Google Shopping search campaign when you want to reach customers searching for the range. Facebook product advertising can suit discovery through catalogue-based creative and audiences. Compare them using the same eligible lamp variants and commercial objective.
| Decision | Standard Shopping example | Facebook catalogue-ad example |
|---|---|---|
| Customer context | Product search and comparison | Discovery or renewed interest while using social content |
| Product inputs | Merchant Center data and product groups | Catalogue items and selected creative formats |
| Eligibility | Shopping ads account and item policies | Meta account, catalogue and advertising requirements |
| Controls to record | Locations, Online filter, included groups, bids and budget | Product selection, audience settings, placements and budget |
| Destination | Advertised colour, price and online delivery | Same colour, price and delivery promise |
| Purchase comparison | Named purchase action and attribution settings | Named purchase event and attribution settings |
Meta’s catalogue-ad training describes sales campaigns with broad-audience and retargeting approaches. It also covers catalogue product videos and other creative formats. Record the actual audience and placement settings before launch, including which automation can expand delivery.
LAMP-SAND could answer an existing lamp search on Google or appear in a social room-styling creative. Use the same order-value definition for both campaigns. Compare attributed orders with store order identifiers, since both platforms can claim credit for the same purchase. Separate click-through and view-through credit when comparing reports.
Both examples are paid advertising. Free Shopping listings aren’t a third paid campaign with a zero-cost bid. Neither platform is universally cheaper or better. Product demand and creative fit affect the result. Contribution after costs decides whether the campaign deserves more budget.
Use Semrush alongside the authoritative campaign records
Use Semrush to investigate observed Shopping listings, then verify a campaign decision in your own records. Its PLA Research documentation describes Positions, Competitors and Copies reports for product listing ads. They cover search terms, competing domains and observed creative.
The PLA Positions guide describes selecting a market database, device and reporting period. It shows product titles, destination URLs, listed prices and positions, with search-volume estimates. Access requires the relevant Semrush Advertising Toolkit entitlement. Use a database covering your intended market. Another country’s observations don’t establish Australian visibility.
A fictional research question is, Which product destinations appear for sand table lamps? Enter a competing domain, select the relevant market and filter Positions by the lamp term. Inspect the listed title, URL and price. Suppose a hypothetical result shows a competing sand lamp linking directly to its variant page.
Treat that result as a lead. Open LAMP-SAND’s actual destination, check its Merchant Center record and inspect its Ads inclusion. Correct a confirmed destination mismatch without changing bids at the same time. Save the observed result and the account evidence separately.
Semrush observations and estimates don’t reveal your actual Ads cost or verified purchase value. An absent listing in a research report doesn’t prove a campaign outage. Close the diagnosis using your included item’s state, subsequent Ads delivery and the matching order records.
Give your store its own shopping agent
Customers ask in their own words. Your agent finds, compares and checks out from your catalogue, right on your site.