![rw-book-cover](https://d34adp677peecb.cloudfront.net/static/images/article4.6bc1851654a0.png) ## Metadata - Author: [[Olga]] - Full Title:: Rethinking How Analytics Gets Done - Issue 340 - Category:: #🗞️Articles - Read date:: [[2026-11-05]] ## Highlights > • [Data suggests that longer trial periods perform better](https://substack.com/redirect/15a420f9-92f3-4674-a3a5-8b063353750a?j=eyJ1IjoiM2kxMHhzIn0.1Kpky8Knu0EyceTU0RV_phz4GkjLwPr5RprErnU4tsk). And yet, teams largely still opt for shorter trials. No one knows why. > • “*A hard paywall protects margins, but it caps your audience at the people who’ll pay right away*.” Upselling and winback perform much worse for apps with hard paywalls. ([View Highlight](https://read.readwise.io/read/01m6kcbhkt887z0y1sebbn35jy)) > According to a recent market intelligence report, search engines generated 162B monthly web visits in June 2026, compared to just 8B for AI chatbots. That’s nearly 20× more traffic. Google alone generated 98B monthly visits, compared to ChatGPT’s 5B - roughly 19× more. ([View Highlight](https://read.readwise.io/read/01m6kcg31bedr6mzgyr70cksr4))