You may have seen articles circulating in recent weeks claiming that it is now proven: "the long wait between seasons is responsible for the drop in viewership!" These articles are based on a study by Luminate in the US, which you can read here if you wish. As always with Luminate, the institute mixes various factors and draws connections between unrelated elements, but let's move past that; this provides us with a good starting point for an analysis to determine if this is truly the case.
Since it’s a very long study, here’s a summary.
Preamble: No series ever “gains” viewership.
I think we should start with the obvious: no series ever "gains" viewership. I have already discussed this topic extensively in an article you can read here, which addresses the viewership of Season 2 for Netflix series recently and serves as a good companion to this analysis.
Why do I say this, for those who haven't read my previous article on the subject? Simply put, in the age of streaming, no one discovers a series at its third episode of Season 2. When you click on a title tile on your streaming service, you invariably start with Episode 1 of Season 1. Thirty or forty years ago, on traditional TV channels, this might have been possible, but that is no longer the case. We all start with Episode 1 of Season 1, which remains the most-watched episode of any series on streaming platforms.
The perception of increased viewership stems from the observation window. If I show you this graph of viewership for the three seasons of Squid Game on Netflix over the first four days, one might argue that Season 3 performed better than Season 2, which performed better than Season 1, ergo: "Squid Game gained viewership season after season!"
But if we zoom out slightly, looking at the first 91 days for example, here is what we get: Squid Game Season 1 achieved higher viewership than Season 2, which in turn outperformed Season 3. There is a significant erosion rate between Seasons 1 and 2, followed by a more moderate decline between Seasons 2 and 3.
The only thing that increased was the fans' enthusiasm for the series during the first few days of its release, growing stronger with each season until the final one. However, in the long term, the most-watched episode of Squid Game remains, and always will be, the first episode of Season 1.
We find this type of analytical error in weekly series within a single season as well. Take this article, for instance, which discusses Lanterns, the latest HBO Max series that allegedly reached its viewership peak with its third episode, attracting 10 million viewers.
What enables this "miracle"? The answer lies in the rest of the article: Episode 1 has actually reached 20 million viewers since its release. Therefore, what Episode 3 achieved was attracting the highest number of viewers over a specific, undisclosed period in Warner Bros. Discovery's counting methodology compared to other episodes. It managed this because, simultaneously, even more viewers discovered Episode 1 and enjoyed it enough to make subsequent episodes a weekly appointment or to catch up the episodes already released. However, 20 million viewers for Episode 1 over three weeks versus 10 million for Episode 3 over a few days still represents a 50% erosion rate between the first and third episodes overall, even if this rate fluctuates afterward. Ultimately, there will always be erosion within a season and between seasons, regardless of whether the wait between them is long or short.
We see this in Netflix's Engagement Reports for long-completed series where all seasons are already available on the platform. Zero weeks of wait between seasons implies no audience loss, right? Well, yes, there is loss for absolutely every multi-season series.
I have selected only a handful for this graph: House of Cards, Orange Is the New Black, 13 Reasons Why, and Unbreakable Kimmy Schmidt. Yet all of them lost viewership over the seasons, despite the next season being just one click away.
Ultimately, this is very logical if you think about it: Episode 1 of Season 1 is the entry point, and every subsequent minute of every episode represents a potential exit door. I suspect that, like me, you have series in your watchlist that you started but never finished, sometimes after two episodes of Season 2 or midway through an episode of Season 3. It is completely random.
So, after this long preamble, I return to my initial observation: no series ever gains viewership after the first episode of the first season. That is simply logical. What we observe is a more or less steep decline in subsequent episodes, fueled by the long-term wave of viewership generated by that first episode of Season 1.
I could therefore dismiss Luminate's study on this point alone, but it raises another underlying issue that is far more interesting to analyze.
Why do viewers become hooked on or drop a series?
What strikes me in this study is that the primary culprit identified for the drop in viewership is the gap between season releases. However, if viewers drop a series during a season or between two seasons, the main culprit seems to lie elsewhere. You know what I am talking about... I am of course referring to the quality of the series itself.
If The Pitt manages to retain its viewers so effectively and minimize erosion, it is not solely due to its weekly release schedule; it is primarily because its quality is such that it keeps the vast majority of its initial viewers captive, prompting them to return in force for the new season. Admittedly, the reduced wait time may play a role in lessening erosion between seasons, but the quality of the series appears to be the main factor to examine here. Yet, this aspect is never addressed as a significant, or even more important, factor compared to the release delay between seasons or the distribution schedule. This is likely because it is an element very difficult to quantify.
How to measure the quality and erosion rate of a series?
"Difficult" does not mean "impossible," which is why I decided to investigate this in my study on the subject: does the quality (or lack thereof) of a series have a greater impact on audience erosion than the release delay between two seasons? To answer this, we would need access to several data points we do not possess, such as viewer satisfaction rates and episode-by-episode erosion rates. Therefore, I had to devise an appropriate methodology, which remains imperfect, and which I will now explain to you:
I used IMDb.com because it is the largest public rating site for films and series in existence, especially since the closure of TV Time. It is global, although it has biases that are difficult to ignore, such as ratings undergoing significant fluctuations when "wokism" is detected by a certain segment of users. The site also appears to over-index Indian users, so it is not necessarily representative of the entire world, but one must start somewhere.
To determine the erosion rate, I compared, for each season of the studied series, the difference in the number of ratings on the site between Episode 1 and Episode 3. I estimate that if viewers drop a series, it occurs primarily between Episodes 1 and 3, although this approach is necessarily imperfect as well.
I also attempted to identify seasons that provide a satisfying conclusion for viewers by comparing the average ratings of the first three episodes against the last three episodes of the season. If a season ends with a lower average rating, after casual viewers have already dropped off and presumably only the more engaged audience remains, one can assume that not all of them will return for the next season. Conversely, if the average rating of the final three episodes is higher than that of the first three, it suggests that the remaining fans are won over and likely to return for the launch of the following season.
I also factored in the overall average rating of these two sets of episodes (a series rising from 6.5 to 7.5 is not quite the same as one going from 8.5 to 9.5, even if the difference between the two averages is identical). This allowed me, through various formulas, to derive a viewer satisfaction index for those who took the trouble to rate the episodes on IMDb.
Next, I naturally accounted for the number of weeks between the releases of different seasons. Regarding the impact on viewership, I focused exclusively on the viewership figures in millions of EVCs (Equivalent Viewers Completed) for the first few days (between 3 and 5) of each season. Over the long term, the data smooths out, as seen in the Squid Game example, but the viewership during those initial few days truly indicates whether fans have returned or not.
Season quality, erosion rate, viewership figures, and wait time: these are the four metrics I will attempt to correlate in the following section.
💶 SOUTENIR LA NEWSLETTER AVEC LES DLCs PAYANTS.
Si vous voulez soutenir la newsletter, vous pouvez acheter l’un des DLCs payants disponibles dans la boutique en ligne sur Patreon. Voici un aperçu non exhaustif de qu’elle contient pour le moment :
Netflix Viewership Data Report : Des récapitulatifs annuels des audiences pour la grande majorité des titres Netflix Original sortis.
US Streaming Viewership Data Report : Des récapitulatifs annuels des audiences en streaming aux Etats-Unis sur les principaux services SVOD.
Netflix Viewership Spotlight : Des récapitulatifs des performances de tous les programmes Netflix Original d’un pays donné.
Merci à celles et ceux qui soutiennent la newsletter à travers ces achats !
From Season 1 to Season 2.
Among the four indicators mentioned, I first chose to retain perceived series quality and erosion rate to categorize the series into four groups:
The Elite: These series seasons demonstrate better audience retention between the first and third episodes and higher perceived quality than the others.
The Comfort Traps: Quality is below average, but retention is better.
The Polarizing Shows: Quality is above average, but retention is lower.
The Danger Zone: Both quality and retention are below average.
Here is where approximately thirty first seasons of Netflix series released in recent years fall within this classification:
Keeping the same color code, I will now plot these series on another axis showing the progression of viewership between the first few days of Season 1 and the first few days of Season 2, alongside the wait time between the two seasons in weeks.
Here is what we get:
The most significant drops occur for series whose first season falls into the Danger Zone (red) or The Polarizing Shows (orange), which seems quite logical. Almost all series in the Elite zone (green) are above average, with the exception of Season 2 of Beef. In my view, the explanation is quite simple: Beef is an anthology series where each season is unique. Fans of Season 1 did not find the characters they appreciated, as that story had a definitive ending. Here, fans are asked to respond more to the concept than to the characters and the story, which seems rather complicated.
The cases of Avatar: The Last Airbender and One Piece are also quite interesting to examine, as they are two adaptations of ultra-famous franchises, presenting a double-edged sword. Either the adaptation appears failed to its fans, as with Avatar, resulting in a sharp decline, or the adaptation is successful, and fans turn up proportionally more for the next season, even if the general public, attracted by the prominence and promotion surrounding the Season 1 release, has likely deserted the ranks.
This also highlights one of IMDb's biases: "geek" titles are rated much more heavily there than young adult or even romantic titles. Consequently, this introduces fairly strong biases, of which I am aware.
The GOAT here is undoubtedly Squid Game, which manages to have the longest delay between the release of its first two seasons and the best viewership increase in its opening days. In this specific case, the long wait worked in the title's favor.
A final interesting comparison in this section is between Nobody Wants This, Running Point, and The 4 Seasons. These are three comedies that released their second season approximately one year after the first, yet they have quite varied fates. These outcomes can be fairly easily attributed to the perceived quality of each series: the one better received in Season 1 had the best retention at the launch of Season 2, while the one with the poorest reception in Season 1 had the weakest retention at the launch of Season 2.
From Season 2 to Season 3.
Beyond Season 1, only the fans remain, which inevitably changes the dynamics. They are not necessarily more lenient regarding the series' quality (the median average remains the same as for Season 1), but they are slightly more loyal, which mechanically raises the median retention rate. Here is what this looks like for Season 2, including the variations compared to Season 1.
Many series slide into the Danger Zone due to the rise in the median retention rate, while Season 2 of Bridgerton demonstrates exceptional retention compared to the others.
This shift in perceived quality inevitably leads to a corresponding shift in viewership for the following season (Season 3), even if the release delay for that season was shorter than it was for Season 2.
NDlA: The Median average is the same for all seasons beyond season 3.
This is notably the case for My Life With the Walter Boys, The Lincoln Lawyer, XO, Kitty, and Squid Game, although in the latter's case, the Season 3 launch remains superior to the Season 2 launch. That is the power of being a phenomenon. The Lincoln Lawyer appears to suffer the most from the mediocre reception of Season 2, even though the release gap between seasons remained unchanged.
Beauty in Black is perhaps the only example here where I believe a rapid release schedule (each season released in two parts spaced a few months apart, significantly reducing the wait) worked in its favor. Season 3 returned with quite exceptional growth in its opening days. However, this could also be explained by the fact that the perceived quality of Season 2 increased significantly compared to Season 1. Fans loved what they saw and returned with even greater force, which likely explains why the series was renewed for a Season 4, despite Season 3 originally being intended as the final one.
From Season 3 to Season 4.
For this transition from Season 3 to Season 4, we will focus on The Lincoln Lawyer, as its third season experienced a genuine rebound in perceived quality among viewers, moving it back into the Elite zone after having passed through the Danger Zone.
This automatically translates into a significant surge in viewership at launch, even though the release delay remains substantially the same as in previous seasons. This is truly the most telling example that the delay matters less than perceived quality. Once again, this seems so logical to me that it is quite staggering to have to conduct such a lengthy analysis to arrive at this conclusion.
We also see that Season 4 of The Witcher dropped significantly at launch, largely due to the recasting of its main character. Outer Banks also declined at the start of its fourth season, but this is also because the first three seasons formed a coherent arc with a beginning, middle, and end. Season 4 attempted to launch a new story, but some fans (myself included) were unwilling to re-engage.
Finally, Virgin River, Emily in Paris, and Bridgerton appear immune to decline upon new season releases, showing very limited increases or decreases. Their perceived quality by the IMDb audience likely differs from that of the general public due to their specific target demographics. It is also worth noting the performance of Sweet Magnolias, whose Season 4 shows a better return than the previous one; however, Season 3 was better received by fans, and this positive momentum naturally carried over to the launch of Season 4.
From Season 4 to Season 5.
As the seasons progress, examples begin to dwindle, but this time we welcome Stranger Things, which quite logically falls into the Elite zone following the release of its fourth season. It is far more complicated for Outer Banks, whose quality index drops significantly in Season 4 (following the death of a beloved character that deeply divided fans). Sweet Magnolias Season 4 was also much less well-received.
Consequently, series that had poorly received seasons show a more or less significant decline, such as the final season of You, Outer Banks, and Sweet Magnolias (which has since been cancelled).
Emily in Paris is a rather peculiar case, I find, as the series fluctuates enormously, independently of its perceived quality by its audience or the delay between seasons. Stranger Things also demonstrates that waiting can work in your favor, but only if a series manages to hook its viewers. Here too, that seems quite logical.
Conclusion
Does a long wait between seasons impact the viewership of a streaming series? Given all the figures mentioned in this analysis and taking into account the inherent biases of IMDb, I believe we can answer: "Yes, but not necessarily in the expected way and certainly not in a predominant way."
The data shows that there is no simple linear correlation where "more wait" equals "fewer views." Series that waited three years saw their launch viewership increase, while series that waited only one year saw their launch viewership collapse.
However, there is clearly a trend-multiplying effect associated with the delay between seasons.
For event series, a long wait is an asset. The delay creates a sense of scarcity and establishes a cultural appointment. The wait does not make viewers forget the series. Instead, it transforms the release into an unmissable event, as was the case for Stranger Things, Squid Game, Wednesday, and to a lesser extent, Bridgerton and Ginny & Georgia.
For "run-of-the-mill" series, which experience average drop-off and decent satisfaction, the wait seems fatal beyond 18 months. Here, the wait acts as a slow poison; these series are not hated, but simply "forgotten." Below 18 months, if the season's quality holds steady, the audience remains stable. This is perhaps why Netflix tends to renew certain series for multiple seasons outright: to ensure they are released more quickly.
Finally, for series in crisis, those with a sharp drop-off, a decline in quality, or an anthology format, the wait delivers the final blow. Even if it is not necessarily the sole cause of the decline, it amplifies the severity of the fall.
There is also a distinction to be made between series set in already known universes, which benefit from a potentially very strong audience pool at the launch of Season 1, and those that are completely original, whose novelty attracts new viewers more strongly over the long term. This is precisely why comparing The Pitt to The Rings of Power, or Squid Game to One Piece, seems somewhat very simplistic.
In short, all this is to say that time does not kill series. It only kills those that failed to create an emotional bond strong enough with their audience to survive oblivion. For the others, time is the best ally to transform a launch into a global phenomenon.
If you enjoyed this issue of Netflix & Chiffres, why not share this newsletter with someone who might also be interested?
Vous pouvez me retrouver ici :
🎙️ Le podcast mensuel Netflixers sur l’actu du streaming.
💬 Mon compte BlueSky.
📧 Par mail, pour toutes vos questions et commentaires.
💶 Sur Patreon, où je vends des dossiers sur les audiences du streaming.
🎞️ Sur Letterboxd, le Serializd des films.
📺 Sur Serializd, le Letterboxd des séries.
