This week's mix is light on genuine news and heavy on platform housekeeping: Google admitted AI Mode ads monetise worse than traditional search, while Google and Microsoft both pushed tutorials and updates around conversion tracking, app measurement and Performance Max controls ahead of the holiday season.
On Google's Q3 earnings call, chief business officer Philipp Schindler said AI Mode queries currently have lower monetisation than traditional search, citing FX headwinds, APAC retail disruption from the EU's de minimis change, and media spend normalising after an unusually strong World Cup quarter. He said Google is expanding ads and multiturn ad formats in AI Mode to more markets and queries.
This contradicts Google's earlier line that AI Overviews ads monetise the same as standard search, though Schindler is careful to say this statement is about AI Mode specifically, not AI Overviews. Search Engine Roundtable's Barry Schwartz reports the earnings call directly, and quotes Glenn Gabe's take on X that Schindler's point is narrow: each AI Mode query makes less revenue right now, but Google expects monetisation to improve as ads expand and as it argues many AI Mode queries are upper funnel research rather than purchase intent. I'd qualify Gabe's framing: upper funnel usage growing faster than monetisation is exactly the scenario that should make advertisers wary of blindly trusting AI Mode traffic quality in reporting, since more queries per session does not automatically mean more commercial intent per query.
Google published an official tutorial walking through App Connect: linking an app SDK such as Firebase, importing in-app conversion actions, switching key actions like purchases or registrations from secondary to primary for bidding, and enabling Smart Bidding on campaigns once those actions are set as primary.
This is a straightforward product walkthrough from Google Ads' own YouTube channel, so the facts are reliable, though the framing that this unlocks 'real ROI' is a vendor claim worth testing against your own numbers before trusting it fully.
A second official Google Ads tutorial explains the Deep Link Setup section of App Connect, covering how to identify deep linking opportunities, fix misconfigured deep links with a validator tool, and build new campaigns from top-performing deep links.
Another direct product tutorial from Google's own channel, so the mechanics are trustworthy. The claim that this 'unlocks full revenue potential' is marketing language from the video description and should be tested rather than assumed.
Microsoft Advertising's October update introduces AI-powered campaign creation that builds a Performance Max campaign from your website and materials in minutes, makes age and gender targeting generally available on Performance Max with device targeting rolling out, opens audience exclusion pilots and refined brand exclusions, adds search term and auction insights reporting with landing page reporting coming mid-November, opens a messaging guidelines pilot to keep AI-generated ad copy on brand, and takes Conversions API generally available for server-to-server conversion data from websites, apps, CRM and point-of-sale systems.
This is Microsoft's own official blog, so the feature list is accurate, though claims about AI 'taking work off your plate' and campaigns being 'set up to perform from day one' are vendor framing that should be judged on results, not promised. The most consequential item here is Conversions API reaching general availability, because richer first-party signals feeding Smart Bidding directly addresses the core problem with automated Performance Max campaigns: bidding quality is only as good as the conversion data behind it. This aligns with my view that PMax is unsafe for lead gen without strong offline conversion quality, so Microsoft giving advertisers a proper server-to-server route to feed that data is a genuine improvement, not just noise.
Google published an official walkthrough of Ads Data Manager showing how to check data source connection status, edit field mappings and update frequency, and crucially attach a conversion goal, since Google Ads will not track a data source's activity until a conversion action is linked to it.
This is a direct, low-risk product tutorial from Google's own channel. The key practical fact worth flagging to clients is the one buried in the transcript: uploading first-party data to Data Manager does nothing on its own until it is explicitly tied to a conversion goal.