On June 15, Google flipped a switch that made every Performance Max product report look better overnight. Impressions up. Clicks up. Conversions up. If you walked into a pipeline review this week with those numbers, you might have gotten a round of applause. You shouldn't have.
The update expanded product-level reporting to include data from all eligible networks, not just Search. That means your reports now capture activity across Video, Display, Discover, Gmail, YouTube, and Maps, where your products were already serving but weren't being counted. The spend was always there. The conversions were always there. Google just wasn't showing them to you in the product view.
This is a measurement change, not a performance improvement. And if your team conflates the two, you'll make bad decisions with real budget consequences.
What Actually Changed
Before June 15, product reporting in Performance Max reflected only Search network activity and Standard Shopping campaigns. If a product ad ran on YouTube or Display, the cost and conversion data existed somewhere in your account, but it didn't roll up into the product-level view. You were flying partially blind.
Google's announcement confirmed that the expanded reporting now covers all Performance Max networks, plus Video, Demand Gen, and App campaigns where products are served through Google Merchant Center. The structure is unified: all campaigns using Merchant Center now return results through a single shopping_performance_view.
The practical effect is a one-time jump in reported metrics. Google warned advertisers to expect this, but warnings don't always reach the people building board decks. If you're comparing June to May without context, you're comparing apples to a fruit salad.
The Historical Comparison Problem
Here's where the math gets uncomfortable. Google explicitly stated there will be no historical backfill. Data before June 15 reflects the old, narrower scope. Data after June 15 reflects the new, broader scope. They are not directly comparable.
This creates a structural break in your time series. Any month-over-month or year-over-year analysis that spans the June 15 cutoff will show an artificial inflection point. Your CAC payback calculations, your channel-mix models, your incrementality tests: all of them need to account for this discontinuity.
The fix isn't complicated, but it requires discipline. Annotate the break in every report. Segment your analysis into pre- and post-June 15 periods. And when someone asks why Performance Max "improved" in mid-June, have the answer ready: it didn't. The measurement did.
Using the New Data Without Getting Burned
The expanded reporting does offer genuine value. For the first time, you can see how individual products perform across Google's full inventory from a single view. Google's documentation confirms you can now segment by channel, including Search, Display, Discover, Gmail, YouTube, Maps, and Search partners.
This visibility matters for product-level decisions. If a SKU is driving conversions on YouTube but underperforming on Search, you now have the data to investigate. If a product category is eating budget on Display with no return, you can see it. The channel performance report, which Google introduced in 2025, now has a cost toggle and ROI columns that make this analysis actionable.
The right move is to use the Network (with search partners) filter to break down where activity is coming from. Build a baseline from June 15 forward. Compare like to like. And resist the temptation to credit the algorithm for gains that are purely a function of expanded measurement.
What This Means for Your Forecast
If you're running Performance Max at scale, this change has implications for how you model future performance. Your historical data now understates the true scope of product activity. Your forward projections, if they're based on pre-June 15 trends, will underestimate what the expanded reporting will show.
The conservative approach is to treat June 15 as a reset point. Build your forecasts from the new baseline. If you need to show continuity for board reporting, create a bridging analysis that explains the measurement change and isolates the underlying performance trend.
This is also a good moment to revisit your attribution assumptions. As industry practitioners have noted, the new channel-level data can be misleading if you analyze each channel in isolation. Performance Max optimizes holistically across channels in real time. A channel that looks inefficient on average ROI might be driving incremental conversions that wouldn't exist otherwise. The marginal contribution matters more than the average.
A Two-Week Pilot Plan
If you haven't already, here's what to do before your next pipeline review:
First, pull your product reports for June 1-14 and June 15-30. Compare the metric totals. The delta is your measurement artifact, not your performance change.
Second, segment your June 15-30 data by network. Identify which channels are now contributing to your product-level metrics that weren't visible before.
Third, update your reporting templates with a clear annotation explaining the June 15 change. Anyone reading the report six months from now needs to understand why the numbers shifted.
The risk here isn't technical. It's organizational. Someone in your company is going to see the June numbers and draw the wrong conclusion. Your job is to make sure that person isn't the one presenting to the board.
