BZFD - BuzzFeed Engaging The Makers Of ChatGPT Is A Good Business Move

Summary

• BuzzFeed has exploded lately with first a Meta partnership that has BuzzFeed on a content creation deal for Meta's platforms, and then an AI announcement.
• Using the OpenAI API does make sense given BZFD's cost structure. For the commodified, mass-market content that BuzzFeed supplies this is perfect.
• The economics of producing AI based content is very good, and with compensation being such a large part of the COGS, a lot of content can be economised upon.
• Still, this is a commodifying industry, and they aren't growing profitably. We pass.

BuzzFeed Overview

BuzzFeed ( BZFD ) is a relatively infamous company, but its fortunes have been very good over the last couple of days. First, it is being engaged to produce content specifically for Meta's ( META ) platforms, and secondly it announced that it'll start using the OpenAI API, OpenAI of course being the makers of the headline ChatGPT MLP model, to the delight of markets. Both of these are very good things for the company and justifies some buying. However, while the Meta deal introduces some less risky revenue, BZFD has an issue doing profitable growth. AI could be used effectively there, but there's nothing terribly interesting about BuzzFeed given latent dilution effects from SBC and the shrinking cash reserves.

Newest Headlines

The newest headlines around BuzzFeed is that Meta is engaging it to create content on Instagram and Facebook. There is a symbiosis between the companies, as BuzzFeed saw most of its growth thanks to reach it go outside its own website, primarily on Facebook with its listicles and oft-derided quizzes. The revenue from the deal amounts to a pretty small proportion of BuzzFeed revenues, maybe around 2.5%, but markets likely see this as a situation where the companies are looking like they're allying and therefore see it as a positive.

The even more recent headline is around the company engaging OpenAI , by presumably paying a license to use its API, to generate AI modeled content. We are actually optimistic about the positive increment this could have on the business. BZFD are already doing pretty big layoffs , and they'll be able to fire many more people who make quizzes and lists that could be generated, even if only partially, with a well-placed prompt. The quizzes alone account for about 4% of their daily produced content, and other content like what articles and some listicles probably can be created in part with AI, of course with supervision. News is harder to replace because there needs to be some fact checking, but certainly for creative content, the latest high parameter language models like ChatGPT are capable of generating convincing text and even achieving humour, and could reduce the writing burden for the current rate of content production. It seems that the majority of BuzzFeed content could become addressed by AI. Since it almost costs nothing to generate pages of content with the OpenAI licensed API, there should be free margin gained on providing more content using AI.

Bottom Line

The cost structure highlights that these efforts may be worthwhile. The majority of the cost structure is on the COGS line, primarily from costs compensating the people producing the content for BuzzFeed.

If the majority of the content could be produced more rapidly, and in some cases ideas could be entirely originated by the AI, there's the potential for sharp revisions in the gross margins. We think that with the current technology, if properly applied, it's possible that $20 million gets added to the bottom line here, if not more.

The issue with BuzzFeed is that they've so far not been able to produce profitable growth. They are also subject to the pretty fickle friend of search engines and social media platforms for achieving reach. It is hard to optimise and there is brutal competition. All it takes is for changes in how platforms curate content and BuzzFeed content could become even lower priority. Finally, there's a lot of share-based compensation, or SBC for short, that could be incoming as the company continues to pay executives in stock. Dilution could be as much as 22% based on the shares reserved for options-based compensation .

The AI angle is interesting, but besides that it's not looking great. Easier to say no to this one.

For further details see: