Google Ads has now put the operating rules for Smart Bidding Exploration in writing for retail campaign types. Advertisers running Shopping or Performance Max with Target ROAS bidding can accept a return between 5% and 30% below their target, allowing Google’s algorithm to bid higher on high-potential queries the campaign already qualifies for but often misses.
That sounds like a simple lever. The fine print, first reported by PPC News Feed on October 22 and credited to Hana Kobzová, makes it a real strategic choice—especially for accounts built on portfolio bid strategies.
How the tolerance actually works
The reduction is relative, not a subtraction of percentage points. If your Target ROAS is 200% and you choose a 10% tolerance, your effective target becomes 180%. At the maximum 30% tolerance, a 200% target falls to 140% and a 500% target falls to 350%.
Google adds an accuracy band: you can expect to land within roughly 15% of that effective target over one to two conversion cycles. PPC Land’s calculation shows that means a 200% target with a 10% tolerance could return around 153% and still sit inside the expected range. With a 30% tolerance, a 200% campaign could feasibly return about 119% and still count as on track.
Google describes targeting as a “fishing pond” and Smart Bidding Exploration as the “rod and reel.” It does not enlarge the pond; it bids more aggressively on long-tail queries already inside it.
Three rules before you switch it on
- Eligibility: Search, Shopping and Performance Max campaigns on Target ROAS bidding qualify. Feedless Performance Max gets exploration on Search traffic only.
- One toggle for feed-based PMax: A single switch covers both Search and Shopping traffic when a Merchant Center feed is present.
- No portfolio strategies for retail: Portfolio bid strategies, in Google Ads or Search Ads 360, work only for Search. Moving Shopping or PMax campaigns into a portfolio automatically turns exploration off.
Budget is the quiet dependency. Google says exploration shouldn’t run on budget-limited campaigns, and hitting a cap may temporarily pause it. There’s no published timeline for when it restarts.
The real choice for retail teams
This is not a free reach upgrade. Exploration asks advertisers to loosen two controls at once: the ROAS target and the budget cap. For many D2C and retail accounts, that lands right before peak trading, when cash preservation and return are already under pressure.
The payoff is selective. Google’s comparison page explains that exploration raises bids on specific traffic segments, while a manual target cut applies everywhere. Internal Search data showed campaigns accepting a 10% lower ROAS averaged 27% more unique converting users and 19% more conversions—but no equivalent Shopping or PMax numbers have been published.
For Shopping campaigns, Google says exploration is fully compatible with AI Max. AI Max broadens the query pool by unlocking text ads; exploration then bids more dynamically inside that broader set. The combination hands both reach and bid intensity to the algorithm, leaving the advertiser with a target and a tolerance.
The practical starting point is a guarded test. Choose a campaign-level Target ROAS retail campaign with an uncapped budget, set a 5–10% tolerance, and hold a control where possible. Measure over one to two conversion cycles, and set your own floor because Google’s expected band is wider than many marketers assume.
Source: PPC Land
