<?xml version="1.0" encoding="UTF-8"?><oembed><type>video</type><version>1.0</version><html>&lt;iframe src=&quot;https://www.loom.com/embed/33585abcb54e4bed81438357667a729b&quot; frameborder=&quot;0&quot; width=&quot;960&quot; height=&quot;720&quot; webkitallowfullscreen mozallowfullscreen allowfullscreen&gt;&lt;/iframe&gt;</html><height>720</height><width>960</width><provider_name>Loom</provider_name><provider_url>https://www.loom.com</provider_url><thumbnail_height>720</thumbnail_height><thumbnail_width>960</thumbnail_width><thumbnail_url>https://cdn.loom.com/sessions/thumbnails/33585abcb54e4bed81438357667a729b-7e991d4a99ed78df.gif</thumbnail_url><duration>145.989</duration><title>2.1 - GrowthStack - Traffic Mix and Reach </title><description>This Loom explains why you should validate experiment audience assumptions before running an A/B test, especially on mobile. The author says to first understand what the traffic really looks like and which browser attributes are actually being sent, using Growth Stack and its coverage line to confirm splits by device and exact dates. They note that only 6% of visits even send a browser value, so a Safari-only test would reach about 3% of visits, around 400 customers, instead of the thousands teams expect. The Loom concludes that small samples can cause bad decisions, so know your traffic and gaps before designing the test.</description></oembed>