Snap Q3 2023 earnings: revenue beat, ad stack improvements yielding fruit

Snap released its Q3 2023 earnings results yesterday:

  • Revenue of $1.19BN represents 5.4% growth on a year-over-year basis and 11.3% sequentially. This was the first quarter since Q4 2022 in which global revenue grew on a year-over-year basis. Snap beat analyst expectations on both EPS (+$0.02 vs. -$0.06 expected) and topline revenue ($1.19BN vs. $1.11BN expected);
  • ARPU declined globally (-5.8%) and in North America (-16%) but grew significantly in Europe (+15.3%) and in the Rest of World reporting region (+10.1%);
  • DAU increased by 12% globally, to more than 400MM;
  • The company didn’t offer “formal” guidance for Q4, citing “the unpredictable nature of war,” although it did note in its investor letter that its internal model forecasts revenue of between $1.32 and $1.38BN for Q4, which would represent year-over-year growth of between 2% and 6%;
  • Snap noted in its earnings press release that COO Jerry Hunter, who ascended to the COO role from SVP of Engineering a little more than a year ago, would depart the company;
  • The company’s stock price whipsawed after the print, rising by nearly 20% at one point to ultimately ending flat.

On the whole, but certainly in the context of Snap’s recent struggles, Snap posted a strong quarter. Management’s comments about its preference to not provide formal Q4 guidance — because of uncertainty related to the war — pose important questions about the volatile nature of brand advertising revenue. Those comments, in full:

In addition, we observed pauses in spending from a large number of primarily brand-oriented advertising campaigns immediately following the onset of the war in the Middle East, and this has been a headwind to revenue quarter-to-date. While some of these campaigns have now resumed, and the impact on our revenue has partially diminished, we continue to observe new pauses and the risk that these pauses could persist or increase in magnitude remains. Due to the unpredictable nature of war, we believe it would be imprudent to provide formal guidance for Q4.

Assuming this is true, the situation highlights how capricious and sensitive to current events a brand-oriented advertising business can be:

  • When tragic, harrowing, or grim events transpire in the world, those events end up being broadcast to social media;
  • But brand advertisers are especially sensitive to having their ads displayed next to the imagery that documents those events;
  • While Snap isn’t entirely dependent on brand advertising, and its content format doesn’t necessarily lend itself to documenting current events, it’s still clearly susceptible to brand pullback when the horrific artifacts of appalling world events end up being discussed or debated on its platform;
  • Twitter is likely the poster child for this risk.

Separately, Snap noted in its investor letter that its subscription tier, Snapchat Plus, grew to more to “more than 5 million subscribers in Q3.” While this is impressive, at $3.99 per month per subscription, Snapchat Plus is delivering something in the range of $60MM in revenue per quarter — or 5% of Snap’s total Q3 revenue. So while Snap’s progress with its subscription product is commendable, it’s not really needle-moving for the business, and it’s unclear what scale it can achieve (for what it’s worth: Snap revealed “more than” 4MM subscribers for Snapchat Plus in Q2).

Snap’s executives highlighted on the analyst call the progress that the company has made in overhauling its advertising infrastructure in the wake of Apple’s ATT privacy policy. The company referenced its 7/0 attribution model (7-day attribution window for conversions attributed to swipes up, with a 0-day window for views), which I explained in more detail in my analysis of Snap’s earnings last quarter, multiple times throughout the call.

From Derek Andersen, Snap’s Chief Financial Officer, on the call (emphasis mine):

We continue to make really significant investments in the ad ranking and optimization, incorporating a much broader range of signals into the ad platform and driving much larger models. And we’ve also instituted a much faster pace of experimentation. All of that’s leading to more precise conversion predictions, improved ROI for advertisers. And as we noted in the letter in particular, 7/0 has been a bright spot in the quarter in terms of driving Pixel Purchase behavior and the year-over-year and quarter-over-quarter growth that we saw in lower funnel…

The reference to lower-funnel here relates to the 7/0 model’s ability to optimize for conversions, and not just clicks or other ad interactions.

Derek Anderson, on the categories in which Snap has seen the most success in improving its direct response advertising product (emphasis mine):

Really what we’re seeing is we’ve been focusing very heavily on getting our go-to-market efforts and where we’re seeing success in the progress of the ad platform really well aligned. And so making sure that we’re selling through the products and the optimizations that are going to work for our advertisers and help them grow their business, that’s what led us to see success with a pretty large variety of sectors. But some of the ones that we’ve talked about recently that I can share with you is we’re seeing good traction with CPG, restaurants and traveler categories that have been doing well on the platform and combinations of e-comm businesses also doing well, and that’s going to align well to both — what I talked about earlier which is the really good progress we’ve seen on 7/0 Pixel Purchase optimizations and what that’s meant for lower-funnel revenue growth year-over-year and quarter-over-quarter. And then also a little bit of the success that we’ve seen with those new brand takeover products and the traction that we got with those in Q3 that was helpful to the sequential improvement in revenue there.

Evan Spiegel, Snap’s CEO, on optimization improvements that have been rolled out (emphasis mine):

I could speak just briefly to the progress we’ve been making on the tech side of things. We’re making a lot of headway with our ML platform, being able to run bigger models that have more features. So I think that — I feel good about our progress there, obviously more to do. I think kind of the atomic unit of our performance business right now is 7/0 purchase optimization. That’s working well for a lot of advertisers. And so, really what we’ve been trying to do is then build on top of that to solve slightly more sophisticated advertiser needs … So, for example, with our app business, we’ve been beta testing event optimization for app advertisers. So, rather than just optimizing around an install, optimizing around something like completing a level or something like that in a game. And those sorts of optimizations are really important for certain performance advertisers. So, I think taking that sort of fundamental building block that’s working with 7/0 and then iterating on top of that to meet more advertiser needs is really important right now … I also think there’s more we can do in terms of advertiser cold start for the long tail of advertisers. It takes maybe a little too long and maybe a bit too much spend for these smaller advertisers to get ramped up and to find conversions quickly. So, we’ve also been iterating a lot there and making progress there. But overall, I feel good about where the various tech pieces are right now.

These are all important features that Snap has fallen significantly behind in bringing to its advertising suite. Meta introduced its App Event Optimization (AEO) bid strategy — which allows app advertisers to bid on in-app events like purchases, rather than merely installs — in 2016 (I walk through the mechanics of AEO in this post). But while Snap lags Google, Meta, and others on in-app event purchase optimization, the upside from here is fairly obvious: the severity of the lag should give some indication of the scale of the opportunity. Obviously, the size of the lag also invites concerns related to execution.

In response to a question related to the sharp ARPU growth in the European reporting region, Derek Anderson (emphasis mine):

Look, I think part of what you’re seeing there is the fundamental improvements that we’re making to the ad platform and the ad products in our go-to-market. So, for example, when you’re looking at the year-over-year growth in the European region, what we’re seeing there is that the fundamental improvements that we’ve made to the DR ad platform … So, some of the things I talked about earlier in terms of the big investments we’ve made in infrastructure to drive ad ranking and optimization, the work that we’ve done to incorporate a broader range of signals and features into those models and then those models themselves becoming much larger. The faster pace of experimentation that we’re seeing on the DR ad platform, all of that’s driving out much more precise conversion predictions. And that’s driving improved ROI for advertisers. And you can see that, as I mentioned earlier, in the improvement in lower funnel year-over-year and quarter-over-quarter.

Later in the answer, Anderson was conspicuous to repeat something that Spiegel had stated earlier in the call, which is that North America was the reporting region with the greatest sequential growth in the quarter, at 14%. Last quarter, concerns arose around the degree to which growth skewed to Snap’s Rest of World region, which is its largest at 211MM DAU in Q3 (52% of total) and its fastest-growing, with a weighted-average sequential growth rate of 5.2% (vs 0.42% for North America). The company seemed intent on this call on focusing attention to North American revenue growth, instead.