The five degrees of separation between a “fact” and a FACT.
This is part II of a totally random series of me being irritated by a media industry statistic. Part I was in 2013. The five year gap is entirely uncorrelated to the number of times I have actually been irritated by media industry statistics.
In a press release yesterday, Parse.ly highlighted some very interesting changes in Google-related traffic referral patterns for news publishers. About 3/4 of the way down the page, the post also noted the apparent value of AMP traffic:
I have been working with AMP pages for two years, and have seen that and similar numbers asserted previously and never bothered to track down the source. But, “2x” seemed surprising given what I have seen in the actual metrics.
Parse.ly provides a November 2017 Moz.com post as their source of the data.
Moz.com links back to the home of AMP, AMPProject.org referencing a graphic in an October 2017 post.
AMP provides a footnote to an actual original source — a Forrester Consulting report dated November 2017. (Obviously it was probably released to Google in October.)
The Forrester report references the “2x” number twice. First in the executive summary.
And — the actual canonical reference on page 5.
So — do AMP pages actually drive 2x the average time-spent-on-page for news publishers? Maybe, maybe not — though not in my experience. But, Google’s paid-for research speaks only to time spent on e-commerce sites.
Everyone in this chain of evidence actually did some due diligence: linking to a source, noting Google paid for the research and in Google’s cases actually footnoting the original document. But, we still have to address the fact that an awful lot of nuance got lost in those five degrees of separation.