Google updated how it measures Branded Searches, the conversion type that counts when someone searches for your brand on Google or YouTube after seeing your ad. Search Engine Land reports the refresh changes the metric's availability, attribution, and reporting, most notably a shift to a 7-day default conversion window, the removal of Performance Max as an eligible campaign type, and a formal placement under the Consideration goal. For B2B SaaS teams that had started reading brand-demand signal out of the metric, these are not cosmetic tweaks.
Branded Searches matters to SaaS because the buying journey is long and indirect. A prospect watches a YouTube ad, does nothing measurable, and three weeks later searches your name and books a demo through what looks like organic branded traffic. Direct-response conversions never credit the ad that planted the intent. Branded Searches is Google's attempt to quantify exactly that hidden step. This guide explains what the metric is, what the August 2026 update changed, why it is reporting-only, and how to fold it into a B2B SaaS measurement stack without letting it distort the numbers your board actually cares about.
What the Branded Searches metric actually measures
Branded Searches counts the event of a person searching for your brand on Google or YouTube after being exposed to your ad. Google launched it in 2025 as an always-on substitute for Search Lift experiments, which required you to design a study, hold out a control group, and wait for results. The always-on version gives you a continuous read: every day, a count of how many ad-exposed users went on to look you up by name. It is an upper-funnel demand signal, categorized as of this update under the Consideration goal, and it sits conceptually between an impression and a conversion — proof that an ad moved someone from passive exposure to active interest.
The distinction from a normal conversion action is the whole point. A trial start or a demo request is a direct-response event the user completes on your site. A branded search is an intent event the user completes on Google's surface, not yours, which is why only Google can measure it. For B2B SaaS, where the gap between first ad exposure and first site conversion can span weeks and multiple stakeholders, that intermediate signal is genuinely useful. It tells you whether awareness spend is manufacturing demand or just burning impressions, a question that pipeline metrics answer far too late. If you already segment brand from generic intent, our guide to brand campaigns for B2B SaaS covers the defensive side of the same coin.
What the August 2026 update changed
Three changes stand out. First, the default conversion window is now 7 days, down from the 30-day view-through window described when the metric first shipped, and you can set it anywhere between 1 and 30 days. A shorter default tightens attribution to searches that follow soon after exposure, which reduces over-crediting but can undercount slow-moving enterprise journeys. Second, Performance Max is no longer an eligible campaign type; the metric now applies to YouTube and Demand Gen only. Third, Google formally categorizes Branded Searches under the Consideration goal and treats it as a primary conversion action that reports in the Results and All conversions columns rather than the standard Conversions column.
The Performance Max removal is the change most likely to break an existing reporting habit. Accounts that ran PMax for awareness and read brand-search lift from it will simply stop seeing the signal there, and will need a YouTube or Demand Gen campaign to recover it. The window change quietly alters your historical trend line too: comparing post-update Branded Searches counts against pre-update ones is not apples-to-apples, because a 7-day window will report fewer branded searches than a 30-day one over the same exposure. Note the shift in your reporting so a lower number is not misread as declining demand. For the general principle of matching a window to your sales cycle, see our piece on conversion windows and the B2B SaaS sales cycle.
Why reporting-only is the right design
Branded Searches is available for reporting but not as a bidding optimization goal, and its data lands in the All conversions column, not the standard Conversions column Smart Bidding chases. This is the most important property of the metric for a SaaS advertiser to understand. Because it is walled off from bidding, turning it on cannot retrain your Target CPA or Target ROAS strategy to optimize for brand searches instead of pipeline. You get full visibility with zero risk of an algorithm deciding that manufacturing cheap branded searches is the way to hit its target.
That matters because a branded search is a textbook example of a proxy that looks great and pays nothing. If bidding could optimize for it, it would find the cheapest possible way to make people type your name — low-intent audiences, broad awareness placements — and your cost per real conversion would quietly climb while a vanity number soared. Keeping the metric in reporting protects you from that failure mode by design. The discipline it demands from you is simple: read it as a diagnostic, keep it out of your primary conversions, and hold your campaigns accountable to the outcomes that appear in the standard Conversions column. Our guide to optimizing for SQLs, not leads makes the same argument one funnel stage deeper.
Where the signal now lives: YouTube and Demand Gen
After the update, Branded Searches is captured on YouTube and Demand Gen campaigns only. That is a coherent scope, because those are the formats built to create demand rather than harvest it. A YouTube pre-roll or a Demand Gen placement is far more likely to make a prospect remember your name and search it later than a bottom-of-funnel search ad, which is already catching people who searched. Confining the metric to these campaign types keeps it honest: it measures demand generation where demand generation actually happens.
For B2B SaaS accounts, the practical implication is that reading brand lift now requires you to actually run YouTube or Demand Gen, not just Performance Max. If your media plan is entirely search and PMax, you will see no Branded Searches data and should not conclude that your ads generate no brand demand — you simply have no campaign type eligible to measure it. Teams that want this signal should carve out a dedicated upper-funnel YouTube or Demand Gen line, size it as an awareness investment, and use Branded Searches as one of the few available reads on whether that spend works. Pair it with a view of the metrics your board cares about so the awareness line has an accountability frame.
How to fold it into a B2B SaaS measurement stack
Treat Branded Searches as an upper-funnel layer that sits above, and never replaces, your pipeline metrics. Keep it in All conversions, exclude it from the primary conversions Smart Bidding optimizes, and chart it as a trend rather than a target. The most useful way to read it is directionally and in context: is your Branded Searches count rising as you scale YouTube spend, and does that rise show up downstream as more branded search volume in Search Console and more brand-driven pipeline a few weeks later? A consistent lead-lag relationship is the evidence that your awareness spend is doing real work.
Set the conversion window to match your category's speed. The 7-day default suits most SaaS, where an interested prospect looks you up within a week or abandons the thread. A high-ACV, committee-driven purchase with a long research phase may justify stretching toward 14 or 30 days to capture slower branded searches, at the cost of looser attribution. Pick one window, document it, and hold it constant so your trend line stays comparable over time. Above all, resist the temptation to promote a rising Branded Searches number as a headline result to leadership — it is a diagnostic that explains why pipeline moves, not a substitute for the pipeline itself. For the broader context this fits into, see our overview of conversion tracking for SaaS.
Pitfalls to avoid
The first trap is comparing counts across the window change. A 7-day default will report a smaller number than the old 30-day view-through window over identical exposure, so a year-over-year or pre/post comparison that ignores the change will read a measurement artifact as a demand decline. Annotate the date of the update in your reporting and treat the series as broken at that point. The second trap is expecting continuity from Performance Max; if your brand-lift reporting depended on PMax, that data path is gone, and its absence is not a signal about performance.
The deeper trap is letting a soft metric drift into a hard target. Branded Searches is easy to grow with cheap, low-intent awareness spend, which is precisely why Google keeps it out of bidding. If you or your agency start reporting it as a primary success metric, you will eventually optimize toward it informally — shifting budget to whatever makes the number climb — and reproduce by hand the failure mode the reporting-only design was built to prevent. Keep it in its lane: a useful window into upper-funnel demand, read alongside, and always subordinate to, the SQL and pipeline numbers that determine whether the account is actually working. If you are auditing which signals deserve weight, our note on attribution window length for B2B SaaS applies the same skepticism to attribution settings.