FB: Recovery? Yes … But Then What?

sagawa
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

SEE LAST PAGE OF THIS REPORT Paul Sagawa / Tejas Raut Dessai

FOR IMPORTANT DISCLOSURES 203.901.1633/.901.1634

psagawa@ / trdessai@ssrllc.com

twitter.jpg @PaulSagawaSSR

April 10, 2018

FB: Recovery? Yes … But Then What?

While much of the public commentary on FB and the Cambridge Analytica affair has been overblown, the resulting pressure on engagement and ad sales is real. FB will sharply curtail 3rd party access to individual user profiles, eliminating a significant benefit to advertisers vs. rivals with more restrictive policies. It may also implement the privacy changes mandated by the EU’s GDPR regulations (e.g. clear opt-ins by users required for ALL commercial use of personal information) across all geographies. We believe the impact of these circumstances for FB sales (-7-10%) and earnings (-10-15%) over the next few quarters could be substantial. Still, we do not expect the effects to be permanent – linear TV is in irreversible decline, there are few other good digital choices for advertisers, and the idea of a lasting user rebellion is improbable. However, there are longer-term concerns. Once the shift from TV plays out, digital ad growth will slow toward the growth of the economy – FB has not developed non-ad monetization strategies. Moreover, FB’s position with younger demographics continues to deteriorate. The current “scandal” diverts management attention from these issues. We do not see read across to GOOGL and other digital platforms – FB’s chief rivals have more conservative 3rd party access policies, targeting ads to categories rather than individuals. Furthermore, GOOGL is aggressively pursuing numerous high-potential non-ad opportunities.

  • FB has been roasted in the press. In 2014, 270K FB users downloaded a personality quiz app, granting its sponsor, Cambridge Analytica (CA), access to their FB profiles via a free API. At the time, 3rd party apps could access not only names, profile photos, and basic demographic information, but also friend lists, enabling CA to build a database of more than 30M users tied back to the psychographic data collected from the quiz answers. While the transfer of the data from the app publisher to CA, and a subsequent failure to delete the data as requested by FB (related to a change to restrict access to friend lists) were both in violation of FB policy agreements, public and political opinion is holding the company responsible for the use of the data in the 2016 election. This leaves FB at considerable near-term risk of weakening engagement, punitive regulation, and softening commitment by advertisers.
  • FB must change. The EU’s GDPR privacy rules will take effect in May, requiring all use of personal information to be explicitly approved via clear, opt-in agreements. These regulations will be a burden to FB, which, unlike most rivals, allows its advertisers to target individuals rather than categories – historically, a lucrative competitive advantage. FB is being pressured to extend its new EU compliant policies to all geographies, likely eliminating direct targeting entirely. This, and the general negative atmosphere around FB, is could dampen ad sales by 7-10% in the coming quarters with commensurate impact on EPS.
  • FB will recover. These pressures will not last forever. FB will establish new, compliant, ways to effectively target advertising by category rather than specific user profile and the glare of publicity will shift to some other company’s crisis. Meanwhile, the deterioration of viewership for linear TV leaves $178B in global TV ad spending ($77B US), some 35% of the total measured media ad market, up for grabs. Advertisers have few other choices beyond FB (and GOOGL, AMZN, TWTR and maybe, SNAP), particularly if new policies leave alternative ad platforms without access to FB or GOOGL for ad targeting. FB, with its nearly 2 billion MAUs, is a unique and valuable platform for users to connect with each other and for advertisers to connect with them. It cannot be replicated.
  • BUT there are serious long-term threats. The erosion of TV ad sales – the last bastion of traditional media – will fuel digital advertising over the next 4-5 years, but at some point, share shifts will equilibrate and digital ad growth will asymptotically approach the 3-4% annual growth in economy-wide ad spending. Unlike its primary rival, GOOGL, FB has not invested to develop non-advertising business opportunities. The $2B deal for VR video game platform Oculus has been a disappointment, and the move to messaging is an ad-driven extension of the primary social media platform. Attempts at adding e-commerce have failed. We also note that FB’s penetration amongst US 12-17-year-olds has fallen to less than 50% – an ominous trend that cast further shadows on longer term growth. In 5-6 years, when the digital ad market is growing with the global economy and the 12-17 segment has aged to 17-24, it is not clear how FB can meet expectations.
  • Caution is advised. We believe that it is too early to jump back into FB. 1Q18 earnings and forward guidance could reveal nasty surprises and spur significant downward revisions. CEO Zuckerberg appears far more intent on atoning for the misuse of the FB platform and remaking its data capture and use policies than in assuaging advertisers or pursuing new monetization strategies. In this context, we fear that the recovery will take more than a quarter or two. There is even a risk that a combination of stalled user engagement and softening advertiser commitment throws the stock into a lengthier TWTR-like purgatory. We would wait for expectations to overreact to the downside.
  • FB’s loss is GOOGL (and TWTR’s) gain. We are less concerned that GOOGL will suffer a similar crisis. GOOGL’s data policies have always been far more conservative then FB. 3rd parties have never been able target to individuals on the platform, they can only target to well categorized groups. While GOOGL obviously collects enormous amounts of data on its users, that data is generally tied to the specific purposes of the service that they are providing – e.g. search queries for Search, location for Maps, genre preferences for YouTube, etc. – a practice that would seem more palatable than using the contents of a user’s messages or their appearance on someone else’s friend list for targeting. We believe GOOGL and TWTR (which also has more conservative data policies) could be beneficiaries of advertiser reticence to engage with FB.
  • The other FANGs are not as dependent on ads. We also note that GOOGL has made significant investments to develop high potential non-ad driven opportunities, such as enterprise hosting/software, eCommerce, and autonomous Transportation-as-a-Service. For AMZN, advertising is growing revenue option for its unmatched eCommerce platform. Other major digital players – AAPL, NFLX, MSFT, etc. – have almost no exposure to advertising or to privacy advocacy.

Mr. Zuckerberg’s Congressional Perp Walk

Facebook CEO Mark Zuckerberg is in a tough spot. Historically, Facebook has been generous to developers, offering a free API that allows apps that obtain appropriate user permissions to access its user profiles to target advertising. In industry circles, this was viewed as a good thing, enabling small apps to compete for ad dollars with the big boys, and pushing Google to keep its platform open to 3rd parties as well. Facebook, with its “Hacker’s Way” ethic and relentless focus on connecting as many people as possible, was not focused on the possibility that nefarious 3rd parties might use this access in ways that its users might find objectionable. Still, Facebook recognized some of this potential, cutting 3rd party access to its users’ friend lists in 2016 and requesting developers who had accessed that data previously to delete it.

During the 2016 elections, the Trump campaign hired Cambridge Analytica, which had retained a list of at least 30M Facebook users linked to psychographic profiles that it had constructed using an app that paid 270K respondents $2-4 to complete a personality test, and then identified their friends’ Facebook accounts via the API (Exhibit 1). Cambridge Analytica used the data to target political ads on Facebook, despite having agreed to delete it. It is not clear that these ads were particularly effective in persuading voters, but the strong emotions around the election have stoked criticism from the media and politicians, in turn, spooking investors who fear a backlash from users and advertisers.

Exh 1: Timeline of FB Data Controversy

The criticism has been harsh and misleading. Facebook has been accused of “selling” user profiles. It doesn’t – rather it allows 3rd parties to link profiles that they build from their own usage data to specific Facebook accounts using a free API (as noted previously, the policy that had allowed these apps to also link to the accounts of each user’s friends was ended two years ago.). Under the current policy, the app ends up with a list of specific Facebook users that can be specifically tied to their activity on their own app and targeted for advertising. This benefits Facebook, as these ads can only be placed on its platform. In contrast, Google, Twitter and other platforms do not enable targeting to the specific user name (Exhibit 2, 3). Rather, ads are targeted to groups of users sharing specific characteristics – the buyer knows the targeting criteria but not the names of the specific accounts targeted.

Exh 2: Third Party Developer and Advertisers Data Access Comparison

Likewise, none of Facebook’s user data was “leaked”. Cambridge Analytica used data collected by a personality test app created by a Cambridge University professor, which was linked to Facebook profiles. Cambridge Analytica did not have its own access to most of the details of those profiles. Still, the perception that Facebook allowed the Trump Campaign to mine its social graph to find likely undecided voters and then target them with ads containing “fake news” designed to sway the election to Trump remains – even though it is demonstrably not true. In this context, Zuckerberg found himself called to testify before the House Committee on Energy and Commerce to answer for Facebook’s uncomfortable place in the budding “scandal”.

Exh 3: Profile of User Data Gathered by Digital Ad Leaders

The Hacker’s Way – Off Course

Zuckerberg initial response to the breaking crisis was measured, drawing further criticism, but more reflective of the cognitive dissonance for the leader of an organization that both tilts politically liberal and deeply feels its mission to connect all people on earth. On one hand, giving wide access to 3rd party apps and providing an open forum for sharing has been a cornerstone of its mission, but being used as a tool to

Exh 4: Data Policy Updates Adopted by FB after CA Scandal

influence elections, particularly on behalf of a candidate that most Facebook employees likely oppose, is spurring deep self-reflection.

Indeed, Zuckerberg’s extended round of mea culpas could keep the issue on the front page for longer than it might have otherwise, and its willingness to consider radical changes to address its responsibility in the matter could hinder the recovery of its ad sales momentum (Exhibit 4). This is exacerbated by the looming May 2018 start of stringent EU privacy regulations that prohibit ad targeting to European citizens based on personal data without explicit opt-in agreements spelled out in clear terms. Facebook is the company most compromised by these rules – users would likely decline to share the more “personal” detailed revealed via their posts to their friends if asked explicitly – and we have already modeled a 7% 1st year drop in EU ad revenues in response (Exhibit 5).

Exh 5: Digital Ad Spending in Western Europe Forecast, 2015 – 2021

Exh 6: Key Changes Brought by European GDPR