ANEW - Outlook For Media Sector As 'Streaming Wars' Becomes 'Streaming Détente'

2023-11-22 16:30:00 ET

Summary

• The AI impact on social media and internet search companies.
• Will established big tech names be disrupted or boosted by AI?
• Can the AI laggards in tech become AI winners?

Transcript

Jim Stillwagon: The core profit centers of legacy media, which are TV affiliate fees and advertising, have been under duress now for several years. Pay TV households peaked at 100 million in the US back in 2012. We're now closer to 70 million and declining, with no floor in sight. Netflix (NFLX) has thrived in an environment where media consumption has shifted from legacy pay TV bundles to streaming services.

But if you look back over the past couple of years, this pandemic to reopening cycle, there are two counterintuitive takeaways. First, COVID lockdowns were actually negative for Netflix in retrospect. And second, the end of the easy capital era has been a major positive. So taking each of those in turn, on the first point, yes, Netflix benefited from a subscriber bump during COVID lockdowns, but COVID also kicked off a wave of competition from the rest of legacy media.

Theaters were closed. Theme parks were shut down. Every studio, almost in unison, decided that they were all in on streaming. Disney (DIS), Warner (WBD), Universal, Paramount (PARA) (PARAA), they all pulled licensed content back from Netflix, launched their own streaming services at uneconomical price points, and then marketed those services so aggressively that subscriber acquisition costs inflated across the industry.

And then, meanwhile, the COVID demand spike actually obscured some problems that were building within Netflix's own business. By their own admission, they should have reined in password sharing sooner-- certainly before they ended up with 100 million unpaid viewers globally. So Netflix was not a pandemic beneficiary. But then, to that second point, why would they be thriving in a higher rate environment?

It's somewhat ironic given that Netflix was burning cash and tapping the high-yield markets for much of the past decade. But they were able to scale to the only profitable pure play streamer globally, with 250 million subscribers now worldwide. By comparison, legacy media is still trying to figure out how to profitably scale a streamer.

Disney+, Hulu, Paramount+, Max-- collectively, those services lost $25 billion over the past three years. That's not sustainable, certainly not in a rising rate environment with pay TV cash flows declining. So Netflix's peer group has effectively retrenched. We've gone from the streaming wars to something closer to a streaming detente.

You look at Warner, for example. They've just decided to license some of their high-profile IP back to Netflix. DC Comics, "Dune," "Band of Brothers"-- all available on Netflix now. Warner needs to address their overlevered balance sheet. Netflix has their checkbook open.

So overall, the media ecosystem, it's in flux. But we think Netflix is well-positioned to capitalize on this competitive rationalization phase of streaming media.

Greg Bonnell: What about the consumer internet space? I mean, the whole reason why I can sit in my living room and stream all these movies on my big TV is the fact that I'm a consumer of the internet. And then artificial intelligence layered into that, what's happening there?

Jim Stillwagon: Well, I would say there have been a handful of major shifts in computing architectures over the past four decades-- mainframes, PCs, mobile, cloud. It's early, but AI has the potential to be that type of paradigm shift. And it's shaping up to be incredibly computationally intensive.

So our first order of priority has been to focus on the picks and shovels plays. But we don't necessarily know which consumer app will be the enduring winner of the AI age. We don't even know if Google's Gemini (GOOG) (GOOGL) model will outperform GPT-4 when it gets released in early '24. But we do know that every contender at this point will be utilizing a lot of high-end GPUs and networking equipment.

So that dynamic has benefited companies like Nvidia (NVDA), like AMD (AMD). But we have to start asking the question, who is buying all of these high-end chips? And how are they going to earn an adequate return? At some point, we need a graceful handoff from AI hype cycle enthusiasm to tangible AI monetization events. And thankfully, when we look at that large-cap consumer internet space, we're seeing that handoff take place in real time.

So two examples. First, AI is driving engagement higher across social media. Think about services that specialize in short-form video like TikTok, like Instagram Reels. They all rely on AI-powered recommendation algorithms. Traditional social media feeds, they sort from a narrow subset of content. It's really your friends' and families' posts, right?

Think about that step change in content recommendation complexity as you go to high-velocity short-form video, where you need to select from billions of pieces of content across the creator economy globally. And we know that short-form video is much more engaging than traditional social feeds in isolation. The average daily user on TikTok is now spending over 90 minutes per day in the app.

Instagram has seen their total engagement rise by 40% since 2020 when they launched their TikTok clone Instagram Reels. And all else equal, more consumption equals more monetizable engagement for social media. The second example would be how AI is reshaping the ad tech stack behind consumer internet. The old targeting approach of individual ID tracking based on your browser or your device--

Greg Bonnell: Follow you across the internet-- that question, right?

Jim Stillwagon: That's right. That's on its way out, in part due to shifting data privacy norms. And what we're seeing instead are the most successful platforms implementing probabilistic, cohort-based targeting and attribution. And they're leveraging AI to deliver privacy-safe campaigns at scale.

Google has Performance Max. Meta (META) has Advantage+. They're all pushing AI deeper into their respective ad tech stacks. The end game here is to abstract away complexity for small- and mid-sized advertisers that want to engage in performance marketing.

Ideally, you'll have advertisers show up to these large platforms. They'll have basic creative elements. They'll have their return thresholds that they want to hit. Then they hand the keys over to Google and Meta to say, execute the campaign for us. And think about what generative AI could mean for the ease of ad creative.

You could be in a situation where first-time marketers are launching full-fledged video campaigns from scratch. If you lower the bar for experimentation, you'll bring more demand to the ad auction. That's actually one of the major lessons of the past decade for Meta. They've gone from a million active advertisers to 11 million-plus today. Smaller social media platforms, Snapchat (SNAP), Pinterest (PINS)-- they'll have less than a million advertisers. A large broadcast network might have a couple of 100 major advertisers. So we think AI could be that next major unlock for ad auction density as Google and Meta make it easier for advertisers to market their products and services online.

Greg Bonnell: When I think about the disruption of Google all those years ago, it's because people suddenly found that they preferred their search engine to everyone else. They found some sort of formula, and then people just said, I'm googling it. That just meant to search. What does it mean for their search business, all this kind of AI?

Jim Stillwagon: It's a great question. I think AI will evolve search to more of an AI-powered personal assistant-- more conversational, certainly more creative than a text box and 10 blue links. But the big debate right now is whether AI disrupts or entrenches Google's incumbency in search.

I think it was fair to say Google was caught off guard last year by how quickly consumers embraced ChatGPT, fastest app ever to a 100 million users-- within two months. And then you had Microsoft (MSFT), which is OpenAI's partner, their largest external investor, talking about how AI might jeopardize Google's search franchise.

And suddenly, we were fielding questions about the durability of Google's moat in search. Would Google cede share to an OpenAI Bing partnership? Would Google be able to offset the cost of LLM-powered queries? Those doubts were being directed at the same company that pioneered so many of the major advances in AI over the past decade.

Whenever you say "ChatGPT," that "T" stands for transformer. That was the neural network architecture that Google invented back in 2017. So from that trough of AI pessimism around Google in early 2023, Google search share has been remarkably stable-- still control about 93% of searches globally. Bing is still stuck at 3%. Consumer search behavior hasn't changed meaningfully.

So investors were rewarded for staying the course with Google. But now is not the time for complacency. I think we have to be focused on chatbot proliferation risk. What does it mean if you have an AI-infused search box on almost every digital surface, whether it's consumer or enterprise?

Look at what Microsoft is doing, embedding Copilots across the Office suite. What could AI mean in terms of monetization risks for search? You have to remember Google spent the past two decades optimizing every single pixel on that search results page. If that UI shifts to more of a chatbot conversational interface, that could be disruptive for Google's search advertising revenue.

So we aim to have a balanced perspective thinking about Google search. Clearly, there are risks to monitor with AI. But there is further rerating potential if Google can keep shifting the narrative from presumed AI laggard to convincing AI winner in 2024 and beyond.

Original Post

For further details see: