Google’s 2025 Performance Max (PMax) Updates

Google has announced major updates to Performance Max (PMax) for 2025, claiming to improve transparency, control, and efficiency for advertisers. However, marketers remain skeptical, arguing that while these changes add surface-level control, PMax’s core issues remain unsolved.

This deep dive examines what’s actually changing, why advertisers are still frustrated, and how to maximize success with the new tools.

What’s Changing in PMax for 2025?

✅ 1. More Advertiser Control

PMax has been criticized for its lack of manual control, forcing marketers to trust Google’s AI blindly. The 2025 update introduces new targeting and exclusion options:

🔹 Campaign-level negative keywords

• Advertisers can now exclude irrelevant search terms across all PMax asset groups.
• Previously, negative keywords were only available via Google reps (a major pain point).

🔹 Demographic exclusions

• Allows advertisers to block low-converting age groups or specific audience segments.
• Ideal for industries with strict demographic targeting needs (e.g., luxury brands, financial services).

🔹 Device targeting

• Marketers can now control ad spend by device (desktop, mobile, tablet).
• Previously, Google automatically optimized across devices, often leading to inefficient mobile traffic.

🔹 Brand exclusions

• Prevents PMax from associating products with irrelevant brands in Shopping campaigns.
• Helpful for e-commerce advertisers who saw wasted spend on unrelated brand searches.

🔹 URL rules

• Gives advertisers some control over landing page targeting.
• Prevents PMax from directing traffic to irrelevant or underperforming pages.

📌 Verdict:
🔹 These updates provide greater manual control, but still don’t allow full customization like traditional search campaigns.
🔹 Negative keywords and device targeting are major wins for budget optimization.

---

📊 2. Improved Reporting & Transparency

One of the biggest complaints about PMax has been its black-box optimization—advertisers couldn’t see search terms, audience insights, or clear performance data. The 2025 updates attempt to address this:

🔹 Search themes “usefulness” indicator

• Google now labels whether a search theme is valuable or low-impact.
• However, Google still doesn’t show full search term reports, limiting transparency.

🔹 AI vs. manual input tracking

• Advertisers can now see whether search terms were AI-suggested or manually inputted.
• This helps analyze whether Google’s AI is truly adding value or just auto-expanding to unnecessary queries.

🔹 More detailed asset group reporting

• Breakdown of performance by time, device, and ad format.
• Allows for better insights into where PMax is over/under-performing.

🔹 Downloadable performance data

• Advertisers can now export PMax data for external analysis in third-party tools (e.g., Google Sheets, Looker Studio).
• Before, Google kept much of this data locked inside the platform.

📌 Verdict:
🔹 While these updates add some clarity, PMax still lacks true transparency—search term reporting remains heavily restricted.
🔹 AI tracking insights could be useful, but only if Google allows detailed analysis.

---

👥 3. Customer Acquisition & Audience Insights

Google is adding new audience segmentation tools to help advertisers track new vs. returning customers.

🔹 New vs. returning customer breakdown

• Advertisers can now see how many conversions come from new vs. existing customers.
• This is a step toward better attribution, but Google still controls the definitions.

🔹 High-value new customer goal

• PMax can now prioritize acquiring high-value new customers rather than optimizing purely for volume.
• However, Google’s AI still decides what “high value” means.

📌 Verdict:
🔹 This update is useful for e-commerce brands focused on customer acquisition, but still lacks full advertiser control.
🔹 Google’s AI continues to dictate what is classified as a “valuable customer,” making true optimization difficult.

Persistent Problems with PMax

Despite these updates, PMax still suffers from major fundamental issues:

1️⃣ Black-box AI remains

• Google still automatically optimizes with limited advertiser input.
• No full search term reporting → advertisers can’t see where their money is going.

2️⃣ Attribution manipulation continues

• PMax often takes credit for conversions that were actually driven by other channels (e.g., brand searches, organic traffic).
• Marketers still struggle to separate real PMax impact from inflated ROAS metrics.

3️⃣ Budget inefficiencies persist

• Even with device targeting & negative keywords, Google still prioritizes spending over efficiency.
• Many marketers see diminishing returns when scaling PMax budgets.

4️⃣ Unpredictable scaling

• Increasing budgets often hurts rather than improves performance.
• Google’s AI shifts spend in ways that aren’t always logical.

Digital Marketing Freelancer / Agency

What Smart Marketers Should Do in 2025

Since PMax will never provide full transparency, advertisers need to take control where they can:

✅ Leverage new controls strategically

• Use negative keywords, demographic exclusions, and device targeting to limit wasted spend.

✅ Verify Google’s data independently

• Use media mix modeling (MMM) and incrementality tests to measure true impact.
• Tools like Measured (for enterprises) or BlueAlpha (for small businesses) can help.

✅ Manually analyze search themes

• Google’s AI isn’t perfect—manually review search themes to ensure relevance.

✅ Compare PMax to other platforms

• Test against Meta, TikTok, and other ad platforms to measure true incremental impact.
• Don’t rely solely on Google’s reported ROAS.

✅ Avoid over-scaling budgets

• Instead of blindly increasing spend, test in smaller increments.
• Watch for performance drop-offs as spend increases.

Comparison Table: Google PMax 2024 vs. 2025 Updates

This table highlights the key differences between Performance Max (PMax) in 2024 vs. 2025, showing what’s improved and what still needs fixing.

Feature	PMax 2024 🔴	PMax 2025 🟢	Verdict ⚖️
Negative Keywords	Only available via Google reps	Now available at campaign level	✅ Huge improvement
Demographic Exclusions	No control over age/gender targeting	Can block low-converting age groups	✅ Better targeting
Device Targeting	Google optimizes across all devices	Advertisers can prioritize desktop, mobile, or tablet	✅ More budget control
Brand Exclusions	No way to prevent association with irrelevant brands	Can exclude brands from product feeds	✅ Helps e-commerce brands
URL Rules for Targeting	Google decides landing pages	Advertisers can restrict URL usage	✅ More relevance
Search Themes Reporting	No clear insights on keyword expansion	AI now labels useful vs. low-value themes	🔶 Still limited transparency
AI vs. Manual Input Tracking	No way to tell if search terms were AI-generated	Now shows if queries came from AI or manual input	🔶 Could be useful, but lacks full details
Performance Reporting by Device & Time	Limited asset group insights	More granular breakdowns by time & device	✅ Better visibility
Downloadable Performance Data	Data locked in Google Ads	Now exportable for external analysis	✅ Long overdue improvement
New vs. Returning Customer Tracking	No clear breakdown of customer acquisition	Separates new & returning customers at the campaign level	✅ Useful for acquisition campaigns
High-Value New Customer Goal	No prioritization for high-value customers	Can optimize for high-value new customers	🔶 Still dependent on Google’s AI
Transparency & AI Control	PMax operates as a “black box”	Still heavily AI-driven with limited control	❌ Core issue remains
Attribution Accuracy	PMax overclaims conversions	Still no clear way to separate real impact from inflated ROAS	❌ Unresolved problem
Budget Efficiency	Google optimizes for spend, not profitability	AI still spends aggressively without full efficiency	❌ Problem persists
Scalability & Performance Stability	Increasing budgets often hurts rather than helps	No major changes—scaling remains unpredictable	❌ Needs improvement

---

Final Verdict: Are the Google PMax 2025 Updates Enough?

✅ What’s Improved?

• More manual control (Negative keywords, device targeting, demographic exclusions).
• Better reporting (AI/manual tracking, search themes, asset group performance).
• Customer insights (New vs. returning breakdown, high-value new customer goal).

❌ What’s Still Broken?

• PMax remains a black box → AI still dictates optimization without full transparency.
• Attribution remains misleading → PMax continues to over-credit conversions.
• Budget inefficiencies → Google still prioritizes spend over profitability.
• Scaling is unpredictable → Increasing budget can tank performance unexpectedly.