Google Ads API Opens Product Reporting on June 15: What Breaks

Any platform lead who has ever shipped an anomaly detector against Google Ads data knows the dance: thresholds tuned for weeks, then one upstream schema change blows them up at 3am. That is the exact shape of what is coming on June 15, 2026. Google is widening product-level reporting through the Ads API, and the data shift will not look like a launch. It will look like an incident.

What Happened

On April 30, 2026, Dora Sun of the Google Ads API Team posted a notice on the Google Ads Developer Blog flagging three coordinated changes to product reporting that go live on June 15, 2026. As PPC Land reported, the changes touch the shopping_product resource, the shopping_performance_view resource, and Performance Max network coverage all at once.

The first shift: cost and conversion metrics will start returning for Video, Demand Gen, and App campaigns inside the shopping_product resource. Until June 15, those campaign types only returned impressions and clicks at the product level. No spend. No conversions. Anyone running product-linked Video or App campaigns has been flying blind on per-product ROI through that endpoint.

The second shift: shopping_performance_view will return results for every campaign type that uses Google Merchant Center. Previously, behaviour was inconsistent across campaign types, which made unified product dashboards painful to build without custom glue.

The third shift, and the one that will trigger pagers, is Performance Max. All networks will start being returned when querying shopping_performance_view for PMax. Previously, only some networks were included. Google explicitly warns this will produce a one-time increase in all product reporting metrics for PMax campaigns. The increase is not new performance. It is data that was already being generated but not exposed through the API.

One detail that matters more than it sounds: there is no backfill. Historical API requests for dates before June 15, 2026, will not contain the expanded data. No beta. No opt-in window. The cutoff is hard.

Technical Anatomy

To understand why this release lands the way it does, look at the trail Google has left through 2025 and 2026. Channel-level reporting for Performance Max first reached advertiser UI in May 2025, splitting delivery across Search, YouTube, Display, Discover, Gmail, and Maps. That visibility took eight months to reach the API, arriving in Google Ads API v23 in January 2026, which replaced the generic MIXED enum with specific channel breakdowns. Crucially, v23 channel data only exists for dates on or after June 1, 2025. That is the same cutoff pattern repeating now at the product level.

Working backwards: v20 shipped on June 4, 2025, introducing campaign-level negative keywords for PMax (previously only available at account level) plus expanded asset group reporting. v22 followed on October 15, 2025, adding reporting segments like ad_using_product_data and ad_using_video, which tell you whether a given ad pulled from Merchant Center feeds or used video assets. v24 landed on April 22, 2026, with CartDataSalesView for item-level product-sold reporting, nine new lead generation conversion type enums spanning GA4 and Firebase, and VTC optimisation support for Demand Gen and App campaigns.

The June 15 change closes a structural gap left across all of that work. Demand Gen runs across YouTube, Discover, and Gmail. App campaigns drive installs and in-app actions across Search, Play, YouTube, and Display. Both have been generating product-linked spend and conversions, but neither was returning that data at the product level. PMax was returning it, but only for a subset of networks.

My take: this is not a feature launch, it's a transparency catch-up. Google is unifying the surface area of product reporting so that GMC-linked campaigns behave the same way through the API regardless of campaign type. That is good for engineering teams long term. It is operationally hostile in the short term, because the same SQL or BigQuery export run on June 16 will return different shapes and bigger numbers than it did on June 14, with no version flag to gate the behaviour.

Who Gets Burned

The teams most exposed are the ones that have done the most mature work. If you have built anomaly detection on top of Google Ads product reporting, your detectors are now liabilities. Production incidents I've seen in iGaming and e-commerce stacks almost always start the same way: a downstream system trusts a metric, the metric source changes silently, and someone gets paged for a spike that is not real. June 15 is a scheduled version of that pattern.

Performance Max accounts will see the loudest noise. The increase in impressions, clicks, cost, and conversions for product reporting will be proportional to how many networks were previously excluded for that account, and Google has not published the breakdown in advance. So you cannot pre-compute the magnitude. You can only flag the date and quarantine alerts around it.

Agencies running client dashboards are the next blast radius. A 30% week-over-week jump in PMax product spend on a Monday morning client report, with no campaign change to explain it, is the kind of artefact that produces emergency calls. Teams I've worked with in performance marketing know that explaining a data-pipeline artefact to a non-technical stakeholder takes longer than the bug itself.

Analytics teams running year-over-year product performance analysis are quietly burned too. Any comparison straddling June 15, 2026, is structurally invalid unless you annotate the cutoff. A query for May 2026 product spend on Demand Gen returns nothing. The same query for July 2026 returns real numbers. That is not a YoY trend, that is a schema change masquerading as growth.

The uncomfortable read: anyone who treats the Google Ads API as a stable contract is going to keep getting hit like this. The platform is in an extended transparency expansion, and every quarter is going to bring another endpoint that starts returning data it did not return before.

Playbook for Performance Marketing

Before June 15, do four things. First, inventory every alerting rule and dashboard that touches shopping_product or shopping_performance_view. If a rule fires on percentage change in cost or conversions for PMax, Video, Demand Gen, or App campaigns, suppress it for a one-week window around the cutoff.

Second, annotate your data warehouse. Add a column or a dimension table flag marking June 15, 2026, as a schema-break boundary for product reporting. Any BI tool comparing pre and post needs to surface that flag in the UI, not bury it in a wiki nobody reads.

Third, brief your stakeholders now, not on June 16. Send a single-page note to clients and internal finance partners explaining that PMax product spend, conversions, and ROAS will appear to jump on that date, and that the jump is reporting completeness, not media inflation. Get the email timestamped before the spike.

Fourth, if you have been making campaign-mix decisions based on product-level ROAS for Video, Demand Gen, or App campaigns, hold those decisions for two to three weeks after June 15 to let real data accumulate. Decisions made on the first week of expanded reporting will be noisy.

One sharp line for the road: the API did not get better overnight, your blind spots did.

Key Takeaways

• Google Ads API expands product reporting on June 15, 2026, adding cost and conversion metrics for Video, Demand Gen, and App campaigns in shopping_product.
• Performance Max shopping_performance_view will return all networks, producing a one-time spike in product-level impressions, clicks, cost, and conversions.
• There is no historical backfill. Queries for dates before June 15, 2026, will not contain the expanded data, breaking YoY product comparisons across the cutoff.
• Anomaly detection and client alerting rules built on Google Ads product reporting need to be suppressed around the cutoff or they will fire false positives.
• This launch sits inside a longer arc of API transparency work spanning v20 through v24, and more schema-shifting expansions should be expected.