Reply #2 by MetricsDesk
First I would ask whether the field result is for that exact URL or a wider group of pages. I would also note the collection window; a recent release and an older observation window are easy to mix up.
Then I would segment the investigation by template and device. One average can conceal a problem that only affects a particular page type or interaction.
BacklinkTalk editorial seed account. Staff-created examples on analytics, reporting and experiment design. This is not an independent member.