Alphabet (GOOGL) – Earnings Review – January 31, 2024
Demand
![](https://www.bbae.com/wp-content/uploads/Google-logo-2.png)
Google beat revenue estimates by 1.2%. Its 14.9% 3-year revenue CAGR compares to 18.4% last quarter and 24.9% 2 quarters ago. Cloud revenue was 1.7% better than expected as well. Google advertising revenue as a whole missed expectations by roughly 0.5%. This is what most on the street are pointing at to justify this evening’s price action. I think that’s a reach and trying to assign meaning to a few % move from a very hot stock.
![](https://www.bbae.com/wp-content/uploads/image-219.png)
![](https://www.bbae.com/wp-content/uploads/image-220.png)
Source: Brad Freeman – SEC Filings, Company Presentations, and Company Press Releases
Margins
Google met GAAP EBIT estimates and beat $1.60 GAAP EPS estimates by $0.04. Google paid a $10.5 billion tax payment on October 16th. Without this payment, free cash flow (FCF) would have been $18.4 billion for a FCF margin of 21.3%. $18.4 billion is about 20% above consensus expectations (smaller sample size of estimates vs. EBIT and EPS is why I use “about” here).
![](https://www.bbae.com/wp-content/uploads/image-221.png)
![](https://www.bbae.com/wp-content/uploads/image-222.png)
Source: Brad Freeman – SEC Filings, Company Presentations, and Company Press Releases
Balance Sheet
- Share count is down 2.5% Y/Y.
- $113.7B in cash & equivalents; $15B in debt.
Guidance
Google doesn’t generally offer concrete forward guidance, and didn’t this quarter either. The team talked about maintaining healthy YouTube growth, cloud growth and cloud margin expansion. It also sees “notably larger” CapEx in 2024 vs. 2023 to support GenAI infrastructure. It told us last quarter that revenue growth in 2024 would be faster than OpEx growth. The lack of updates here makes me assume that’s still the plan.
Call & Release Highlights
GenerativeAI Overall and Search Specifically:
Like for Microsoft, this was a key theme on the call. Google has a series of foundational models under the “Gemini” umbrella that it’s using across its entire product suite. “Gemini Ultra” is its latest iteration, which is “coming soon.” It’s not yet monetizing it as meaningfully as Microsoft has, but it is enjoying “more AI contribution” in its results. To ensure customers can build whatever they want in a maintained, stable Google environment, it offers “Vertex AI.” Vertex AI is Google’s environment for GenAI app creation from 3rd party developers and its own. Usage of Vertex’s application programming interfaces (APIs) rose 600% Y/Y.
Search Generative Experience (SGE) and Bard are two search-based examples of where GenAI is being used. The products are somewhat similar, but Bard is more conversational while SGE mainly uses GenAI to enhance search result quality. Both are using Gemini; SGE has already cut query latency by a full 40% with it. Google’s team was asked a few times about GenAI search competition and how the landscape is shifting. It dismissed these questions consistently and talked up excitement for the future opportunity.
GenerativeAI and Advertising:
In advertising, Google is using Gemini to accelerate campaign content creation with things like “Auto Created Assets.” This allows ad buyers to surface more relevant results for potential customers with more granular segmentation to improve conversion. Somewhat relatedly, it’s also using Gemini in its Performance Max (PMax) campaign service. PMax is a campaign creation tool that aggregates Google and 3rd party impressions for advertisers to choose from. This has been shown to raise conversion for advertisers by 18% thanks to the larger roster of relevant impressions to select. For things like image discovery campaigns, PMax is raising conversion by another 6%.
During the holiday season, in which retail was the standout segment for Google ad demand, PMax delivered a 60% boost to return on ad spend (ROAS) for a “U.S. big-box retailer.” It raised conversion by 15% for a “global fashion brand” thanks to an omni-channel activation to extend the campaign into buy online, pickup in store.
Google is now working on a new conversational experience within Google Ads (think Bard for campaign creation). It’s in beta testing and is expected to enhance campaign performance and so demand for Google’s advertising tools.
Notably, from a macro point of view, consumers remain in thrifty and deal-hunting mode.
Subscriptions:
Google’s roster of subscriptions now represents a $15 billion business annually. It’s also growing quite nicely with Google One (more storage and tools) crossing 100 million subscribers; both YouTube TV Premium and YouTube Music continue to grow as well. It again spoke on how pleased it is with year one of NFL Sunday Ticket rights.
“Google One is doing incredibly well.” – CEO Sundar Pichai
Google Cloud:
First and foremost, the sequential acceleration in growth, the revenue beat vs. expectations and the resumption in sequential operating leverage were all notable highlights here. That cannot be overstated.
Google Cloud deepened McDonald’s and Verizon relationships during the quarter with AI product add-ons. AI model-building powerhouses like Anthropic are also using Google Cloud’s AI hyper-computer to train its own models. It’s supposedly enjoying key model training cost benefits from doing so.
YouTube:
- Its YouTube create app is “putting production studios in the palm of a consumer’s hand” with GenAI. This app can now translate videos into several languages and automate content creation and delivery as well.
- Google again told us that YouTube Shorts boasts 2 billion users and 70 billion daily views. These metrics are unchanged from last quarter. I would’ve loved to see continued growth in views.
- Google debuted a “less disruptive” ad format in Shorts this quarter, which is part of its plan to continue closing the monetization gap for Shorts vs. traditional content.
- It also debuted an interactive ad format. With it, consumers can engage with TV-based ads on their phones to more easily convert eyeballs into revenue.
Belt Tightening:
As part of its ongoing cost cutting work, which includes layoffs, real estate footprint shrinkage and more, Google continues to make moves. This quarter, it combined Nest, Pixel and FitBit device teams into a consolidated “devices team.” What a concept. This should lead to lower headcount needs, streamlined work and better efficiency. It also wound down lower priority projects during the period.
CFO Ruth Porat is “pleased with the progress” Google is making in “re-engineering the cost base.” It will continue to very selectively hire only engineering and other technical talent. It will also continue to hunt for middle management layers to remove from the organizational structure — just like Meta did. The two companies continue to mimic each other’s CapEx, data center and cost cutting moves.
Still, CapEx will ramp for Google in the coming quarters to support infrastructure investments needed to be a core GenAI player for years to come. That means near term cash flow headwinds, but it is working on data center layouts and other ideas to improve technical efficiency.
More Interesting Nuggets:
- Waymo passed 1 million fully automated trips.
- Accenture is training 90,000 consultants on Google GenAI services. The two are teaming up to build the “GenAI center of excellence.”
Take
I was candidly surprised by Google’s reaction to this report. It was likely just a matter of a very hot stock delivering a rock-solid, but not blow-out quarter. I truly don’t think there are any red flags here to point to, and don’t think bulls should be feeling anxious about this report at all. The FCF hit is purely based on a one-time payment and the cloud outperformance was encouraging. Even as long term investors, it’s never fun to see a stock sell off after earnings. In my non-shareholder opinion, I think shareholders should exhale, zoom out, and think of the investment case as firmly intact. Feel free to disagree, as always. The main item, beyond cloud performance, that would change my opinion is deteriorating search market share via new entrants like Perplexity and ChatGPT. I haven’t seen any material evidence of that happening.
![](https://www.bbae.com/wp-content/uploads/BBAE-Social.png)