Blog

Google turned automatic text overlays on Demand Gen videos on by default. Where the setting lives, when it helps, when it ruins the frame, and how to audit.

A tagging standard you can adopt as written: how GCLID and UTMs divide the work, which values to lock down, and how to migrate without losing historical data.

How to build a seed list that produces lift, what the 2.5 / 5 / 10% reach tiers really mean, and why the similarity slider became a suggestion in 2026.

Engagement halved but revenue flat? The 2026 Google Ads metric changes, plus a 15-minute test that separates a redefinition from a real performance drop.

The branded searches metric in Google Ads 2026: what it counts, where the columns hide, how brand mapping works and why it stays reporting-only.

How YouTube reservation buying works in 2026: fixed CPM, Premier reach, YouTube Select lineups, Masthead, hold windows and when reserved inventory beats the auction.

How Performance Max for travel goals works in 2026: property assets and place IDs, running without Hotel Center, bidding, long booking lag and cancellation handling.

Google admits its filters do not stop invalid leads. Three form checks, a scoring model and the conversion architecture that stops training bids on junk.

Asset limits, listing group tree rules, splitting patterns that work in retail, and the mistakes that make a Performance Max campaign compete with itself.

Since October 2026, Conversion Lift runs self-serve on Search and Performance Max. Thresholds, holdback sizing, study duration and how to read iCPA and iROAS.