Can Social Media Predict Which Catalog Films Are About to Trend?

Can social media actually predict which catalog films are about to trend?

Yes. Not as a forecast or a guess, but as a measurable pattern. When a catalog film starts generating new fan-made clips on TikTok, YouTube, and Instagram, and when the comments under those clips include people asking "what film is this?" and "where can I watch it?", that activity tends to arrive weeks or months before the same title shows up in Google Trends or Wikipedia traffic.

We have now studied this pattern across multiple catalog titles and the result is consistent. The social signal moves first. The traditional metrics follow. The gap between the two is the window in which acquisition and programming decisions can be made early.

What does this look like in real data?

We recently ran a detailed study on Mulholland Drive (2001), the David Lynch film approaching its 25th anniversary in October 2026. We tracked every fan-made clip we could find across TikTok, YouTube, and Instagram. Not posts from studios or press outlets. Just fans.

The numbers: 526 fan-made posts. 61 million views across all three platforms. 149 viewer asks of the "what film is this?" or "where can I watch it?" kind. 2,283 comments expressing love, rewatching intent, or first-time discovery. And 561 comments calling it the best David Lynch film or the best film of the century.

This is a 25-year-old film with a massive, active, young, global online audience. The activity is not a spike. It has been building steadily for over a year, with a clear acceleration in 2026 as the anniversary approaches.

How does the social signal compare to Google Trends and Wikipedia?

This is the most important part of the data. Over the six months from January to June 2026, the number of new fan posts appearing each month rose 202% compared to the same period a year earlier. In a typical recent month, those posts generated about 2.6 million views.

Over the same period, monthly Wikipedia page views for Mulholland Drive fell 20%. Google Trends for the film in Ireland stayed flat. The traditional metrics say nothing is happening. The social data says the opposite.

The reason is structural. Google and Wikipedia only count people who already know the title and search for it by name. The people commenting "what film is this?" under a TikTok clip cannot appear in those counts, because they do not know what to search for. They are invisible to every traditional metric. They are visible to us.

Where is the audience for Mulholland Drive right now?

TikTok is the engine. 256 fan posts, 42 million views, and 99 of the 149 viewer asks. This is where the film goes viral and where new viewers discover it for the first time.

YouTube is the film-fan layer. 196 posts, 15 million views. This is where the longer video essays, scene analyses, and explanation videos live. The audience here already knows the film and is going deeper.

Instagram is where institutions post. 74 posts, 3.5 million views. Criterion, museums, and accounts tied to the 4K restoration. The audience here is smaller but tends to be industry-adjacent.

Geographically, clips are posted from at least 36 countries. The US and UK account for the largest share of creators, but 134 clips come from the rest of the world. Two Irish creators account for nearly 4 million views between them.

What kind of demand signals appear in the comments?

We classified every comment by intent. The four categories: love (people expressing admiration, rewatching intent, or first-time awe), asks (people who do not know the title or cannot find where to watch it), answers (people telling other commenters the film's name), and everything else (opinions, quotes, jokes, plot discussion).

149 comments were direct asks. Of those, 98 did not know the title at all. 51 knew the title but could not find where to stream it. Both are demand. The first kind is invisible to every search-based metric. The second kind tells you there is a distribution gap.

On top of the 149 asks, 245 people answered them by naming the film. That means for every person asking, roughly 1.6 people are answering. The community is doing its own discovery work. But the viewer who scrolls past without commenting, the one who liked the "what film is this?" comment instead of writing their own, is the quiet demand behind the loud question. A single ask with 288 likes is not 1 person wanting the film. It is closer to 289.

Is this a one-time spike or sustained interest?

Sustained. This is the most important distinction for anyone making a catalog decision. A spike means a title is trending for a week. Sustained interest means a title is a library asset with compounding value.

Mulholland Drive has fan posts dating back to 2008 on YouTube. On TikTok and Instagram, the earliest posts in our dataset are from April 2021. The film has generated new fan content every single month for over two years. When David Lynch died in January 2025, every metric rose simultaneously. After the spike faded, the social activity settled at a higher level than before and has been climbing since.

About 30 new fan posts appear every month in 2026. The velocity is increasing, not decreasing. The October 2026 anniversary is the next obvious catalyst, and the audience is building toward it without any marketing or studio push.

What does this mean for streaming and acquisition teams?

Three things.

First, the audience is already there. A programmer or acquisitions team considering Mulholland Drive for a season, a re-release, or a streaming push does not need to create an audience from scratch. There are 61 million views worth of existing attention and 149 people actively asking where to watch it. The promotional work is partly done.

Second, the timing window is open. The 25th anniversary is October 2026. The social build is already underway. A decision made now, three months before the anniversary, lands in the sweet spot where the audience is growing but the market has not yet priced in the moment. A decision made in September lands in a crowded window where every other programmer has noticed.

Third, the data shows which related titles the same audience cares about. In the comments on Mulholland Drive clips, viewers mention Twin Peaks 193 times, Lost Highway 102 times, Blue Velvet 101 times, Inland Empire 48 times, and Eraserhead 33 times. That is a season-building guide written by the audience itself.

How does this compare to other catalog titles you have studied?

We have now run this analysis on multiple catalog titles. In a previous study on a 2019 film, we found the social signal arrived 11 weeks before the Google Trends curve moved. In the Mulholland Drive study, the pattern is even clearer: social posts up 202% year over year while Wikipedia is down 20% and Google is flat.

The gap is consistent. Social curiosity moves first. Traditional metrics follow. The size of the gap varies by title, but the direction has been the same in every case we have studied. We expect to publish more of these studies as the dataset grows.

Can any team do this analysis themselves?

In theory, yes. In practice, three things make it very difficult at scale.

What is catalog intelligence and how does it relate to this?

Catalog intelligence is the practice of identifying which back-catalog films and TV titles are heating up on social media before they appear in traditional demand or viewership data. It is the category we are building at Trace.

The Mulholland Drive study is one example of what catalog intelligence looks like in practice. The goal is not to tell a programmer which film to choose. The goal is to show them the evidence on which to make that choice, earlier than they could see it through Google Trends, Wikipedia, or any other traditional signal.

Trace is building the catalog intelligence platform. To learn more, visit traceclips.com.