Over the last few months, my team and I have been poring over four years’ worth of Netflix top-10-weekly viewership data, two-and-a-half years of What We Watched reports, 10-Qs as well as content spend estimates and title-level metadata licensed from Ampere Analysis. And we also built our own model to categorize shows as serialized, episodic, or anthology.
To benchmark how Netflix converts content dollars into Viewing Hours, we use a raw average, “Cost Per Thousand Viewing Hours.” Think of it as the flipside of Revenue Per Thousand Viewing Hours, a measure of how Netflix turns its engagement engine into revenue.
Netflix itself says: “We measure the impact of our originals on our ability to acquire new members and engagement, which is correlated with retention of existing members. We also seek reasonable economics relative to other exclusive content on a cost per hour viewed. We also take into account critical acclaim and awards for our originals and the impact original series may have on enhancing our brand....” (Emphasis mine).
When I first introduced the CPMH metric, I described its limitations (“Netflix spent $73 to get a thousand viewing hours. If the above sentence makes you cringe, read on.”)