Every week at SeekandHit, we gather for our Weekly Digest to break down the latest updates in digital marketing. In the last session the SeekandHit performance team dove deep into Google’s latest AI targeting features, Meta’s hidden product insights, and the shifting landscape of U.S. search behavior.
1. Is Your Account Ready for Google AI Max?
Google’s AI Max (often referred to as PMax or its latest evolution) is a powerful tool for finding new conversions, but it can be a “money pit” if enabled prematurely. During our meeting, Petra Babic shared a checklist for readiness:
Accurate Tracking: Your conversion tracking must be deduplicated and focused on high-value business outcomes.
Conversion Volume: AI Max generally needs at least 30 conversions per month to work effectively. At 100+, it becomes significantly more stable.
Broad Match Success: You should first prove success with broad match in traditional search campaigns before letting the AI take the wheel.
URL Expansion Guardrails: When using URL expansion, ensure you exclude low-value pages like Help files, FAQs, and blogs.
2. Breaking the Meta “Black Box”
One of the most exciting topics discussed was unlocking product-level data within Meta’s Advantage+ and Dynamic Product Ads (DPA). Currently, Meta doesn’t show which specific products are being shown or clicked within a single ad unit.
We discussed a workflow to solve this:
API Integration: Pulling product_id, impressions, and clicks from the Meta Insights API into BigQuery.
Visualization: Creating scatter charts in Looker Studio to categorize products into four groups:
Star Performers: High impressions, high clicks.
Promising Products: Low impressions, high CTR.
Window Shoppers: High impressions, low clicks.
Low Priority: Low on both.
The Final Data Layer: A step further involves merging these Meta API insights with raw GA4 data. By extracting Product IDs, purchases, and total revenue from GA4, the team can map platform-specific engagement directly to bottom-line results. This integration is the definitive way to arrive at “final data,” providing a complete picture of the ROI for every individual item in the catalog.
This transparency allows our performance specialists to create custom product sets and optimize feeds based on actual business value rather than just platform-level engagement.
Source: Unlocking Meta’s Product-Level Ad Data (Search Engine Journal)
3. Google Demand Gen: More Shoppable & Measurable
Google is pushing Demand Gen to be a full-funnel performance engine. Two key updates stand out:
Shoppable CTV: Users watching YouTube on Connected TV can now browse and buy products directly using their remote control.
Attributed Branded Searches: This new metric helps us justify upper-funnel video spend by showing how many people searched for the brand after seeing a Demand Gen ad.
4. Quick Hits & Platform News
TikTok US Deal Finalized: The long-standing regulatory uncertainty has ended with a deal spinoff to a consortium led by Oracle, Silver Lake, and MGX. ByteDance retains just under 20% ownership. This removes a massive planning hurdle for our clients. (Source)
Search Volume Drop: Reports show Google searches per U.S. user fell nearly 20% YoY. AI is likely cutting down on repeat searches by giving instant answers. (Source)
Centralized Experiment Center: Google Ads has launched a unified hub for A/B tests and lift studies, making it easier to measure campaign impact before scaling. (Source)
Summary
As we move further into the “AI-first” era, our role at SeekandHit is shifting from manual management to strategic oversight. Whether it’s setting proper guardrails for AI Max or using data science to peak into Meta’s black box, the focus remains on data integrity and full-funnel measurement.
Stay tuned for next week’s digest!
Disclaimer: This blog post and accompanying podcast have been fully produced with Gemini AI, and reviewed by a human being, following weekly news-digest meetings at SeekandHit.
For more information contact [email protected].
