Facebook/Meta is the Biggest AI company globally. Surprising but True!
Meta released their quarterly results a few days back. The results were better than analysts’ expectations so the stock popped 12%.
Good for Zuck and Meta shareholders.
But what piqued my interest was the fact that the whole earning call was mainly about Zuck convincing the investors that AI remains Meta’s singular focus FOR NOW.
63% of Zuck’s speech was about JUST AI while 15% was on metaverse and remaining was on overall business.
And Zuck’s AI statements were not just forward looking futuristic statements. What Meta has been able to accomplish for its user engagement and advertiser ROI through AI is commendable to say the least.
There are a lot of companies doing phenomenal and potentially life altering work using AI (all in the future though) - But what Meta has been able to achieve through inducting AI in their current business is equally path breaking and I doubt if anyone else has been able to reap such benefits from AI.
At the least, nobody else is devoting their whole earnings call to AI so in their mind - they have not accomplished much through AI.
Lets start with the dissection of their earnings call and quarterly results:
Overall top of the funnel numbers remain really really STRONG
NUMBER OF USERS USING META APPS
The company boasts of 3 Billion Daily Active Users across all the Apps (3.8 Bn used one of the apps at least once in a month - MAU). Facebook alone sees 2 Billion Daily Active users.
To put this in perspective, these are DAUs for some of the other largest apps:
Tiktok: 600 Mn | Twitter: 230 Mn | Wechat: 450 Mn | Youtube - 122 Mn (Looks really low compared to their MAU of 2.6 Billion meaning that users are not coming to YT daily).
So, the user numbers that Meta is clocking is really really high. It’s like almost all of the internet connected world logs into Meta apps on a daily basis.
However, Meta is seeing plateaued user numbers
Essentially, all of the internet connected world is already using Meta apps (Facebook/Insta/Whatsapp/FB Messenger) daily so what else can Zuck do?
DAU across all the apps has grown from 2.72 Bn to 3.02 Bn in last 8 quarters which is a growth of 1.3% growth per quarter. Essentially, there is not much which can be done here. These numbers would grow as more people get connected to internet and Meta apps are the first apps that they use.
However, for sustaining these DAUs - Meta has had to rely on reels format which has been detrimental to its ad revenue (More on this later in this article).
Now let’s come back to what Zuck said during the earning conference.
As per Zuck, Meta is investing in two major technology waves
Mark’s favourite Tech wave but Universe is not aligning there - METAVERSE
Universe’s favourite and now Zuck also pushing more resources behind the same - AI
AI Wave
Meta has been using AI for two major use cases - Recommendation/Ranking of content and for driving results for advertisers already.
AI powers feeds and reels within FB, Instagram. In addition, it is the magic which drives ads ROI. Discovery engine, reels and ads all rely significantly on the recommendation infrastructure for its phenomenal performance.
Improved engagement in Apps through AI: 40% of the content that we see in Instagram is driven by AI and AI has increased the time spent in Instagram app by over 24%.
ROI on ads is improving: AI based targeting of users within Meta ecosystem has improved the RoI for advertisers leading to stronger demand from advertisers. In fact, Meta ads on open audiences (relying on AI to reach the right set of audiences) consistently outperform targeted ads.
This means that AI is working well on identifying the right set of audiences for ads.
In the future, Zuck is looking to double down on these use cases + also looking at creating new product experiences using generative foundational AI models. These would drive new products and experiences but frankly, Meta does not have much clue on the same (or so it seems basis the communication in the earnings call).
Metaverse
Reality Labs/Metaverse unit revenue was ~300 Mn$ (down 50% from last quarter) on account of lower Quest 2 Sales. The cost of running Reality labs, though, was 4.3 Bn$. The business unit, thus, incurred a loss of 4 Bn$ during the quarter.
Zuck’s Metaverse dream seems to be fading away and is really really expensive for Meta shareholders.
Reasons for Huge Layoffs at Meta
The above graph is singlehandedly responsible for all of the layoffs happening at Meta. Company’s quarterly profits have more than halved from ~10 Bn$ to ~5 Bn$.
Simply put, Revenue seems to have topped out at 30 Bn USD per quarter in the current scheme of things while company’s costs have been ballooning on account of big teams working on 1) Reality Labs 2) R&D for innovation in experiences.
Meta has already laid off 20K employees in two layoff rounds and 1 more layoff round is expected in May. After these layoffs, the company is expected to have 70K odd employees.
But why is advertising revenue stagnant?
During Q1 2023 - Total number of ad impressions increased by 26% but the Price per ad impression reduced by 17%. Growth in ad impressions was mainly from the regions which have lower monetization like Asia Pacific which reduced the cost per ad impression.
US/Canada monetize at ~50 USD per quarter per user (~200 USD per year which is more than Netflix Subscription cost) while Asia and Rest of World (outside Europe) monetizes at ~USD 4 per quarter (5-10% of US monetization numbers).
Reel Monetisation remains a challenge for Meta
Majority of engagement growth at Meta has been delivered by Reels (which Meta has picked up directly from Tiktok).
However, reel monetization has been really challenging for the company. For Per unit time spent in the app - Reel monetization is much lower compared to Feed and Stories as Reels are longer in duration and the opportunities to show ads is limited.
The company needs to figure out the correct ad formats for reels so that the monetization efficiency increases without harming the engagement and user experience
Biggest Bets for Meta in the near future
Automation of advertiser experience
AI based advertiser experience (Advantage+ shopping) is doing well for Meta. The company is planning to introduce AI based Visual creation tools for Insta and Facebook posts and ads.
Idea is to use AI to discover the right audience for the ad and to initially optimise the creative through AI and eventually AI controlling the whole experience including creative.
End goal is for the advertiser to just tell the objective and system doing everything.
Continued improvement in user engagement in reels and better monetisation of reels
Meta is looking for leveraging AI further to help users discover the best reels to increase time in app and is also looking to improve reel monetisation.
Messaging ads are growing well for Meta
Click to message ads have reached a 10 Bn $ annual revenue run rate and are scaling well. Currently, the messaging ads need a lot of intervention from advertiser staff and reporting is also unstructured. The product, thus, is used by mainly small advertisers. Meta is planning to automate all of this so that the product can be used at scale by bigger advertisers as well.
Metaverse
Zuck continues to bet on Metaverse in spite of tepid response from the market. He is spending 4Bn$ per quarter on the same (half of Meta’s overall profit).
New user experiences through AI
Zuck mentioned working on AI based Chat experiences in Messenger and Whatsapp This is another long bet but nothing concrete as of now.
SUMMARY
Whatever is happening - Zuck has been able to get Meta out of the hole. The stock has climbed from its low at ~100$ to almost ~250$ in a span of just 6 months.
AS THEY SAY - NEVER BET AGAINST ZUCK!!