The problem with Snap

Yesterday, Snap’s CEO, Evan Spiegel, wrote a letter to staff declaring that the company would likely see revenue and EBITDA in Q2 fall lower than the range provided in earnings just last month. From the letter, published by The Verge:

Like many companies, we continue to face rising inflation and interest rates, supply chain shortages and labor disruptions, platform policy changes, the impact of the war in Ukraine, and more…Today we filed an 8-K, sharing that the macro environment has deteriorated further and faster than we anticipated when we issued our quarterly guidance last month. As a result, while our revenue continues to grow year-over-year, it is growing more slowly than we expected at this time. We believe it is now likely that we will report revenue and adjusted EBITDA below the low end of the guidance range we provided for this quarter.

The letter also indicated that the company would hire fewer people than previously planned and would look to shed other non-critical expenses. Snap’s stock price dropped by more than 30% when the updated guidance was published; based what Snap reported in earnings on April 21st (the below Tweet is updated in a reply to reflect the correct date), Snap’s growth was suddenly recognized as being roughly half as what was indicated just one month ago.

So what happened?

In his letter, Spiegel cites, “interest rates, supply chain shortages and labor disruptions, platform policy changes, the impact of the war in Ukraine, and more” as current headwinds to the business that contribute to this re-assessment of the company’s fortunes this quarter. Certainly, all of these frictions are real: Meta recently introduced an expansive hiring freeze, as have Coinbase, Twitter, Wayfair, and countless others. And one can’t scroll Twitter for more than a few seconds without seeing advice to startups from venture capitalists on navigating the contracting economy.

But were macroeconomic factors aggravated so acutely and severely in the month since Snap originally issued its Q2 guidance that the company will now miss that guidance?

It’s possible that macroeconomic factors account for the entirety of Snap’s commercial frictions in Q2 2022. But Spiegel’s letter cites ‘platform policy changes’ as an obstacle to the company’s growth. Could latent shocks from Apple’s App Tracking Transparency (ATT) policy have contributed to the company’s slower growth this quarter?

At the recent JPMorgan Global Technology, Media and Communications Conference, Spiegel was asked about Snap’s ability to fully recover lost signal on the platform, and whether the updated Q2 guidance (which Spiegel had first revealed in that discussion) is related to ATT. Spiegel responded:

We’ve certainly made a lot of progress with our first-party measurement solutions, it’s something I think that the entire industry is grappling with. And we may never fully recover all signal…We take opt-in users who have opted into receiving personalized advertisements and allowed us to collect more information about the way that they use our service, and we use that to model what is happening with the opt-out audience. And we offer solutions, like deeper integrations, like the Conversions API and the Snap Pixel to help advertisers better measure the return on investment of their advertising campaigns. While we’ve made those solutions available, we’re still working on driving adoption, and I think the bigger challenge is just trying to build that familiarity and trust with these new solutions. It’s a big change in terms of the way that advertisers have been thinking about measuring their return on investment over the last 10 years.

Later in the discussion, Spiegel states unequivocally that Snap does not collect IP addresses from opted out users. He also states that he believes that obfuscating the IP address is a “likely next step” with respect to future platform privacy policy changes.

The problem with Snap is that it’s not Meta: Snap operates an ad platform in the same hub-and-spoke model as does Meta, but Snap, to my mind, wasn’t the primary target of ATT, and so Snap dodged much of the resiliency scrutiny of the early period after ATT’s introduction. But reporting by the FT alleged a form of probabilistic attribution being undertaken by Snap that may not be compliant with ATT. And no one is in a position to litigate that, as I argue here: not advertisers, and not the non-Meta ad platforms, which are operating in somewhat of a fog concerning compliance (my thesis here is: since Meta knows that ATT was designed to disrupt its platform model, then everything it does to retain signal is by definition not compliant with ATT. Other platforms have less certainty, which is a double-edged sword). I have questioned the lack of enforcement of ATT’s prohibition of fingerprinting, but the reality is that Apple operates in a manner that is inscrutable from the outside.

So what disrupted Snap’s trajectory for the quarter? I see two plausible explanations:

  • Precisely as the company claims, macroeconomic factors caused growth to decelerate rapidly. As I stated in a recent thread, advertisers can (and should!) react quickly to changing economic conditions and examine their ad spend and interrogate their ROAS models rigorously;
  • Snap was forced to alter its modeling and/or measurement infrastructure, either to improve its precision or in anticipation of future privacy changes from Apple. Recall that Meta, in response to the announcement of ATT, initially claimed that it would simply not poll the IDFA on users’ devices and would therefore not be required to even expose the ATT prompt in its apps, which would have left its app-to-web campaigns (for eg. e-commerce advertisers) unperturbed. Just before Christmas 2020, Meta revised that guidance and revealed that it would indeed moderate app-to-web campaigns by the restrictions of ATT. I am told that this change in policy resulted from a sudden awareness on the part of Meta of the universality of ATT’s applicability.

These explanations are not mutually exclusive: both could have happened in concert. Certainly the macroeconomic environment is worrisome, potentially perilous. But I do believe there are further shoes to drop with ATT: whether through a specific mechanism for the enforcement of ATT’s fingerprinting ban, as I hypothesized in How Apple might break fingerprinting in iOS 16, or through a more straightforward form of enforcement. No, ATT doesn’t account for all of the public markets turmoil that has been witnessed across consumer tech over the past few months. But it accounts for some of it, as will future tremors, which are inevitable.