Microsoft Advertising is rolling out a new setting within Performance Max campaigns that allows advertisers to exclude specific audience ad placements. The feature - named 'Excluded content terms' within Audience Ads settings - gives advertisers the ability to prevent their ads appearing in environments they consider unsuitable or off-brand.
On the surface, this looks like a minor quality-of-life update. But it points to something that anyone running Performance Max campaigns on either Microsoft or Google will recognise: automated campaign types make decisions about where your ads appear, and those decisions are not always ones you would have made yourself.
The Placement Problem With Automated Campaigns
Performance Max was built around the premise that the algorithm knows best. You supply assets, set a conversion goal, choose a bidding strategy, and the platform distributes spend across its available inventory. That inventory can include search, shopping, display, video, and - critically - audience network placements that sit well outside the environments you might consciously choose to advertise in.
The practical consequence is that brand safety becomes a reactive concern rather than a proactive one. You discover a problematic placement after budget has already been spent there. Exclusions after the fact help, but they do not recover wasted spend or undo reputational exposure. A setting that lets you define exclusions upfront - before the campaign starts distributing - is therefore more valuable than it might appear.
This is a pattern across both major paid search platforms. Google has progressively added placement reporting and exclusion capabilities to PMax, responding to sustained advertiser pressure. Microsoft is now following a similar trajectory. The direction of travel is clear: advertisers want more visibility and more control, and the platforms are - slowly - providing it.
What This Means in Practice for PMax Management
If you are running Performance Max campaigns and not actively reviewing placement data, you are likely funding placements you have never consciously approved. The audience network component in particular - which covers third-party sites and apps rather than owned Google or Microsoft properties - tends to produce lower-quality engagement at a lower cost per impression. That combination can look efficient in aggregate reporting while quietly dragging down lead quality.
The right approach is to treat placement exclusions as a standing hygiene task, not a one-off fix. For Microsoft campaigns, the new 'Excluded content terms' setting should be reviewed during campaign setup, not added retroactively when something goes wrong. Build a standard exclusion list based on content categories that are irrelevant or harmful to your brand - news and commentary, click-heavy entertainment content, low-quality app environments - and apply it from day one.
For lead generation specifically, this matters more than it does for e-commerce. A retail brand absorbing some low-quality display impressions may still see a purchase conversion somewhere downstream. A B2B advertiser paying for clicks from irrelevant audience placements is generating noise in its conversion data, inflating cost per lead, and potentially polluting the signals that Smart Bidding relies on to optimise the campaign.
The Connection Between Placement Control and Conversion Signal Quality
This is the element that often gets missed in discussions about PMax controls. Every click that comes through an unsuitable placement and fails to convert is a negative signal being fed back into the bidding algorithm. If your PMax campaign is pulling a meaningful share of spend from audience network placements that your target audience simply does not use, the campaign is training itself on bad data.
Restricting those placements does not just reduce wasted spend in isolation - it improves the quality of the signal pool that Smart Bidding is working from. Fewer irrelevant conversions (or non-conversions) in the data means the algorithm has a clearer picture of what genuine intent actually looks like for your product or service. That tends to improve bid efficiency over time, not just in the short term.
This is why placement exclusions belong in the same conversation as conversion tracking quality, audience list hygiene, and asset relevance. They are all inputs into the same system. PMax is only as smart as the environment you create for it to operate in.
How the Platforms Compare on Advertiser Controls
Google and Microsoft have taken similar architectural approaches to Performance Max, but the control features have not developed in lockstep. Google introduced placement reporting and brand exclusion capabilities for PMax in stages over 2023 and 2024, following significant pressure from agencies and advertisers who felt the campaign type was operating as a black box. The ability to exclude specific URLs and content categories came incrementally, not by design.
Microsoft's new placement exclusion feature for audience ads within PMax suggests the platform has absorbed those lessons. Offering content term exclusions as a named setting within the campaign structure is more explicit and accessible than burying exclusions in a separate placement report that many advertisers never review.
For advertisers running campaigns across both platforms, this serves as a useful reminder to audit controls on both sides. The existence of an exclusion feature on one platform does not mean it is configured equivalently on the other. Check what exclusions are active in your Google PMax campaigns, verify whether account-level placement exclusion lists are actually being inherited by PMax, and now do the same for Microsoft.
The Broader Signal: Automation With Accountability
There is a pattern worth recognising here. When automated campaign types launched, the implicit contract was that the algorithm would outperform manual management if you gave it enough room to operate. That is broadly true at scale with strong conversion data. But it created a secondary problem: advertisers lost visibility and control in ways that created real commercial risk.
The response from both Google and Microsoft has been to layer controls back on top of automation. Placement exclusions, brand safety lists, asset-level reporting, search term insights - these are all mechanisms that restore some degree of human oversight to campaigns that are otherwise running autonomously. The feature set is still incomplete by most practitioners' standards, but the direction is right.
The practical takeaway is not to wait for the platforms to build perfect transparency into their automated campaign types. Use the controls that exist now, build exclusion and brand safety lists as a standard part of campaign setup, and review placement data on a regular cadence. The advertisers who treat placement management as an active discipline rather than a one-time configuration will consistently outperform those who leave it entirely to the algorithm.