May 31, 2025

You can implement the advanced modeling with cookieless pings. But it won’t work. This is because of the threshold. For conversion modeling these thresholds are 700 ad clicks over a 7 day period, per country and domain grouping.

New Insights from Ahrefs: AI Overviews in Google Search Ahrefs analyzed 55.8 million AI Overviews (AIOs) across 590 million Google searches, revealing: – AIOs appear in 9.46% of desktop keywords, rising to 16% in the US.– They account for over 54% of search volume impressions, indicating prominence in high-traffic queries.– Top 50 domains capture nearly

Agentic AI can change the shopper journey Dramatically! Below an example of EMARKETER.

New 2025 Google Ads Benchmarks are out from WordStream by LocaliQ Here are the key insights based on 16,000+ campaigns (Apr ’24 – Mar ’25): – Average CTR: 6.66%, but over half of industries saw a decline– Average CPC: $5.26, with 87% of industries seeing increases– Average CVR: 7.52%, with 65% of industries improving– Average

New test/update in the Ads Transparency Center: it now shows “Advertiser identified by Google” followed by “Paid for by [company name]”. Spotted by Anthony Higman.

B2B brands grow by linking their message to real buying situations called Category Entry Points (CEPs). These moments trigger a purchase (e.g. “We need new software after hiring 10 people”). Each extra CEP linked to your brand cuts churn risk. One study found that one additional CEP reduced defection by 5%. Example: If your brand

New Google Ads conversions now default to using the value from GA4 and create a key event (e.g. conversion_event_add_event) that you need to track manually. Spotted by Luca Bartolini and Lorenzo Pugi.

So how do they segment them? Sundar Swaminathan broke it down: — Start with lifecycle stages— Add RFM (Recency, Frequency, Monetary)— Build MECE segments (no overlap, full coverage)— Personalize messaging by segment A customer on their 2nd trip is very different from one who took 20 rides last month. Uber knows this and acts on

The Code Interpreter for Conversational Analytics is now enabled by default when the Gemini in Looker and Trusted Tester features are active for the Google Cloud project linked to a Looker Studio Pro subscription. This preview feature translates natural language questions into Python code to deliver advanced analysis and visualizations. Users with the appropriate permissions

A study by the Ehrenberg-Bass Institute, based on 1,281 elements across 13 FMCG categories in 19 countries, reveals clear insights into which brand assets are most likely to stand out. The top three assets in terms of average consumer ownership are: – Character– Logo– Typography These assets are more likely to be strongly linked to