On June 15, 2026, Google announced that Smart Bidding Exploration — previously a Search-only feature — is now globally available for all Performance Max campaigns without a product feed. For B2B SaaS advertisers running lead-gen PMax campaigns, which almost universally fall into the non-feed category, this changes what the bidding algorithm can do inside your account without you changing a single campaign setting.
The announcement arrived bundled with two other bidding changes — Promotion Mode and the August 17 bidding target optimization — and the Exploration expansion is easy to read past. It should not be. Unlike the August 17 change, which applies automatically, Smart Bidding Exploration is opt-in. But the fact that you now have access to it for every lead-gen PMax campaign means the decision of whether to use it is one every SaaS account manager now has to make deliberately rather than by default.
What Smart Bidding Exploration actually does
The core mechanism is query expansion with a built-in efficiency floor. Standard Smart Bidding optimizes bids on the basis of historical conversion data — it assigns higher bids to queries where users who clicked have historically converted, and lower bids or no bids to queries it has not seen convert. The limitation of that model is that it is self-reinforcing: a query the system has never tested never generates conversion data, so it is never bid on, so it never generates conversion data. Campaigns converge toward the same proven query set and stop finding new conversion territory.
Smart Bidding Exploration breaks that loop by allowing the algorithm to bid on queries outside its established conversion patterns, funded by a ROAS tolerance you define. As Search Engine Land reported when Google announced the expansion, "campaigns using it have seen on average an 18% increase in unique converting search query categories and 19% more conversions." Those figures come from Google's own measurement and represent accounts where Exploration successfully identified queries that converted outside the original established set.
How the ROAS tolerance mechanism works
The tolerance setting is the risk dial. When you enable Smart Bidding Exploration, you specify how much temporary efficiency reduction the system is allowed to absorb in exchange for the opportunity to test unfamiliar queries. If your campaign runs on a Target ROAS of 400% and you set a 20% tolerance, the system can bid on exploratory queries at expected returns as low as roughly 320% ROAS. Queries that accumulate conversions at that tolerance level graduate into the campaign's proven conversion history and are bid on normally thereafter. Queries that do not convert are abandoned and the tolerance budget is spent without compounding — it does not permanently change your blended performance.
The mechanics for Target CPA campaigns work analogously: the tolerance defines how much above your stated CPA the system can spend per conversion while exploring. For a $500 Target CPA with a 20% tolerance, exploratory bids can reach approximately $600 per conversion. Once a query proves itself at that expanded ceiling it is treated as part of the standard conversion set, and bids normalize. The tolerance is not a permanent relaxation of your targets — it is a funded exploration budget with a self-limiting structure. If exploration consistently fails, the algorithm learns that and reduces exploratory bidding without you having to intervene.
What the Performance Max expansion changes for B2B SaaS
Before June 2026, Smart Bidding Exploration was only available on Search campaigns. The expansion to all Performance Max campaigns without product feeds — which covers essentially every lead-gen PMax setup in B2B SaaS — means the feature can now apply across the PMax inventory mix: Search, Display, YouTube, Discover, Gmail, and Maps within a single campaign. That is a different scope than Search-only exploration.
The distinction matters for B2B SaaS accounts because PMax inventory is already broader than a standard Search campaign, and exploration in that environment can expand reach across channels simultaneously rather than just into unexplored search queries. The 42 Agency benchmark data shows Search campaigns currently outperform Performance Max at 553% average ROAS versus 436% for PMax in B2B accounts, which suggests PMax already operates at a wider efficiency spread than Search. Adding query exploration on top of that spread means the performance range within a PMax campaign widens further during the exploration phase. This is a reason to calibrate your tolerance setting conservatively when starting with PMax exploration — tighter than you might for a Search-only campaign.
When to enable it — and when not to
Smart Bidding Exploration is most likely to produce incremental returns in accounts that have already saturated their primary keyword set. If your PMax campaign consistently spends its full budget, hits its conversion target, and has not found meaningful new query volume in the last 90 days, Exploration is a structured way to fund the next layer of discovery without permanently accepting lower efficiency. The 18%/19% Google-reported averages suggest the upside is real in well-data'd accounts, not just theoretical.
The accounts where Exploration adds risk rather than value are those still in or near learning mode. Performance Max requires a meaningful conversion history — generally cited as 50 or more conversions per month in the primary conversion action — to make reliable predictions about which queries will convert. Below that threshold, the algorithm is still establishing its conversion baseline, and exploration adds query variance on top of an already uncertain foundation. The risk is not that Exploration permanently breaks a low-data campaign; it is that the tolerance budget gets spent on inconclusive signals, and the spend is real while the insight is not. If your account is volume-limited, resolve that first via the fundamentals in our Performance Max for SaaS guide before adding exploration on top.
Conversion data quality is the other prerequisite that gets under-weighted. Smart Bidding Exploration works by testing which unfamiliar queries convert — a determination that depends entirely on what the campaign is counting as a conversion. If your primary conversion action is a form fill and form fills do not reliably correlate with pipeline, the exploration algorithm will faithfully discover more queries that generate form fills, not more queries that generate SQLs. The right setup before enabling Exploration is one where your primary conversion action is already a quality signal — ideally a CRM-confirmed MQL or a SQL-importing offline conversion event. This is the same reason offline conversion imports matter for standard Smart Bidding, and the logic applies with greater force to exploration, where the algorithm is actively investing in untested territory based on what it believes a conversion looks like.
How Exploration interacts with the August 17 bidding change
The August 17 bidding target optimization change applies automatically to budget-limited campaigns using Target CPA or Target ROAS. The change makes those campaigns deliver more consistently toward their stated target rather than over- or under-delivering. If your PMax campaign is both budget-limited and on a target-based strategy, and you additionally enable Smart Bidding Exploration, the two mechanisms stack: the system explores new queries within your tolerance floor while also being steered toward your stated target more consistently.
The practical risk of stacking is that a loose or outdated target becomes more expensive in both dimensions simultaneously. The August 17 change uses any headroom in your target; the Exploration tolerance adds further efficiency floor relaxation on top. If your Target CPA was set generously and has not been reviewed, enabling Exploration before tightening the target means the system has two sources of efficiency relaxation available at once. The safe sequencing is to address your targets during the July 6 to August 17 notification window, as described in our post on the August 2026 bidding target change, and only then consider enabling Exploration for PMax campaigns where the target reflects your actual CAC ceiling.
Monitoring Smart Bidding Exploration after you enable it
Google provides the Search terms report in PMax (with limited visibility compared to Search campaigns) and the Insights tab, which surfaces the query categories Exploration is finding. The practical monitoring workflow is to check the Insights tab weekly during the first four to six weeks of Exploration being active, looking for whether the new query categories being discovered align with your ICP. If the exploration is surfacing informational queries or branded queries for adjacent tools, the tolerance may be funding category-level awareness impressions rather than bottom-funnel discovery. Segment by conversion action in the performance data to distinguish between form fills and higher-quality signals.
The metric to watch most closely is not total conversions but cost per SQL or cost per pipeline-stage event if you have that data feeding back via offline conversions. Total conversions can increase while SQL rate drops if Exploration finds queries that generate lower-intent form fills. For B2B SaaS accounts where the sales cycle is long enough that a SQL only closes months after the first click, monitoring at the form-fill level is an unreliable leading indicator — which is precisely why the offline conversion import infrastructure described in our conversion tracking guide for SaaS matters here. If the exploration is working, you should see new query categories accumulating pipeline events over time, not just form fill volume. If they do not accumulate pipeline events within 60 to 90 days, those categories are unlikely to hold up and the algorithm will eventually deprioritize them anyway.
The decision framework for B2B SaaS PMax accounts
Enabling Smart Bidding Exploration for your PMax campaigns is worth doing if your account meets three conditions: it is generating more than 50 conversions per month in a primary conversion action that correlates with real pipeline, it has been spending its PMax budget consistently, and the existing query set has not expanded meaningfully in the last 60 to 90 days. If all three are true, the feature gives you a structured way to fund additional discovery that the standard optimization would not attempt on its own.
If any of those conditions are missing, the right path is to address the underlying gap first. A conversion-limited account should focus on campaign structure and conversion action quality before adding exploration variance. A budget-under-spending account likely has a targeting or bidding constraint that exploration will not fix. And an account whose query set is still growing organically does not need to fund exploration artificially — the standard algorithm is already discovering new territory within its established confidence range. The feature is valuable, but its value is gated on account maturity in a way that makes it unsuitable as a default-on setting for B2B SaaS accounts that have not yet built stable conversion data. If you are unsure whether your account meets the threshold, a structured review — including how your PMax strategy compares to Search — is a good starting point. Our analysis of whether B2B SaaS should run Performance Max covers the structural decision that should precede any bidding feature configuration.