<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Giovanni Perri — Writing</title><description>Notes on analytics, lifecycle automation and experimentation from real client work.</description><link>https://giovanniperri.com/</link><language>en-gb</language><item><title>Running an experiment programme that actually compounds</title><link>https://giovanniperri.com/blog/experiments-that-compound/</link><guid isPermaLink="true">https://giovanniperri.com/blog/experiments-that-compound/</guid><description>Most testing programmes produce a folder of inconclusive results. The difference is not cleverer tests — it is what happens before and after each one.</description><pubDate>Sat, 19 Sep 2026 00:00:00 GMT</pubDate><category>Experimentation</category></item><item><title>Six checks that tell you whether your GA4 is lying</title><link>https://giovanniperri.com/blog/is-your-ga4-lying/</link><guid isPermaLink="true">https://giovanniperri.com/blog/is-your-ga4-lying/</guid><description>An afternoon of work that usually explains the number nobody in the company can reconcile. None of these needs a developer to start.</description><pubDate>Thu, 17 Sep 2026 00:00:00 GMT</pubDate><category>Analytics</category></item><item><title>What I set up first, and in what order</title><link>https://giovanniperri.com/blog/what-i-set-up-first/</link><guid isPermaLink="true">https://giovanniperri.com/blog/what-i-set-up-first/</guid><description>Most growth work fails in the first fortnight, before anyone has run a campaign. Here is the sequence I use, and why the order is not negotiable.</description><pubDate>Tue, 15 Sep 2026 00:00:00 GMT</pubDate><category>Foundations</category></item></channel></rss>