Netflix reveals viewing data globally – what will and won’t this tell us?

#Netflix is to begin revealing viewing data for titles on a global basis, and on a six-monthly cadence. The first data drop covers the period between Jan-Jun 2023, and will contain #streaming viewing hours for roughly 18,000 shows and movies, accounting for 99% of member viewing in the period.

Netflix has long promised ever-increasing transparency on its viewing metrics. There is some speculation that as the push into the ad market continues, viewing data will be increasingly revealed by streamers to support sales efforts. But Netflix was clear that the move is not aimed at the ad market. Ampere’s tracking indicates that ad-tier subscribers still represent a small share of Netflix’s subscriber base, and their viewing patterns may of course differ from the global average. More importantly for viewing metrics providers like Nielsen, Netflix noted that data it provides is not designed as a replacement for third party measurement and that the #advertising market would still require independent solutions to evaluate ad buying strategies.

The metric looks very much designed to support Netflix's global content ambitions

Instead, the metric looks very much designed to support Netflix’s global #content ambitions. While Netflix already provides more detail on individual titles to their creators, it also wants #producers to develop new projects and in particular focus on titles which resonate across its global subscriber base. These are the pieces of content which are most efficient for its business. Providing a global consumption metric helps content creators look beyond their own slate, and develop shows and movies that resonate not just for single Netflix markets, but across the wider global subscriber base.

What does it deliver:

1. The dataset will give global visibility on the hours consumed on Netflix by members – with the titles included providing a near-complete view of the hours streamed over 6 months.
2. Far more detail on the longer tail of shows and movies on Netflix, and additional transparency around which titles are (or aren’t) driving viewing hours.
3. How different categories of titles perform globally – e.g. Netflix noted that Non-English titles already account for ~30% of viewing, and 45% of viewing of English-language titles rely on subtitles or dubbing.
4. Visibility on the significance of licensed vs original content. Netflix revealed that about 55% of viewing is from Netflix Originals and 45% from licensed shows.?

What doesn’t it deliver:

1. Audience sizes – the metrics focus on viewing hours, not breadth of engagement. How many households or people watched a show is a different matter.
2. Local viewing – Netflix was extremely clear that local viewing data will not be revealed, as this represents far too much data to release to competitors.
3. Daily, weekly, quarterly, or monthly data – viewing patterns on a more granular basis won’t form part of this dataset.
4. Episode drop-offs – while there will be some inferences which can be made on viewing time relative to show runtime, it won’t explicitly illustrate whether a show’s audience is sticky, or whether they drift off over the course of the episodes or seasons.
5. Person X also liked – the metrics included will be consumption figures, but won’t tell content creators what similar shows resonated for each individual.

A carefully-judged move by Netflix designed to help suppliers make decisions around creating content which works globally

All-in-all, it will be a fascinating dataset to dive into – and I can guarantee that the Ampere team will be eagerly digesting as soon as possible. It isn’t a panacea to the streaming viewing question though, and it’s very unlikely to herald a new era of open data for the streaming market, but it can be viewed a carefully-judged move by Netflix designed to help suppliers make decisions, to drive efficiencies in the commissioning process, and ultimately create content which resonates globally – supporting ?Netflix subscriber acquisition and retention.