Twitter and the quality of brand advertising revenue

The melodrama of Twitter’s acquisition by a consortium of investors, led by Elon Musk, brings to the fore the commercial contrast between brand advertising revenue and direct response advertising revenue. In The perilous mythology of Brand Marketing for digital products, I attempt to delineate the distinction between brand advertising and direct response advertising:

From an advertiser’s perspective, these different tactics can be utilized in various configurations: it’s possible that one or more could be employed simultaneously or that one might be utilized above the others as the company’s marketing strategy evolves, as Airbnb has done in jettisoning Search advertising in favor of a brand-oriented marketing program.

In the context of Twitter, it’s helpful to consider the quality of brand advertising revenue from an ad platform’s perspective. Last Friday — the day on which Twitter reduced its staff by roughly 50% — Elon Musk revealed that Twitter had experienced a “massive drop in revenue” as advertisers paused their ad spend on the platform. The WSJ had reported earlier in the week that various large agency holding companies had recommended to their clients that ad spend on Twitter be reconsidered, given uncertainty around the platform’s standards related to content moderation as well as the company’s ability to enforce those standards with a diminished workforce.

When a company deploys advertising dollars in pursuit of associative resonance, as brand advertising seeks to achieve, then it’s perfectly understandable that it would want those associations to be positive. Sandwiching a brand ad between pieces of broadly objectionable content could obviously cause damage to consumer perception of the brand that might cascade into weakened sales. It’s perfectly logical that a company would cut brand-oriented advertising spend on a platform with an unpredictable, inconsistent, or fluctuating content moderation policy given the risks inherent in pairing advertising with user-generated content.

Note that I’m not accusing Twitter of changing its content moderation policy or of now tolerating speech that it previously didn’t. Twitter vehemently contends that its content moderation capabilities and standards remain unchanged despite its new ownership and smaller workforce. I see no reason to doubt this. And it’s possible that the advertisers currently pausing spend on Twitter don’t doubt it, either.

Ultimately, it doesn’t matter: brands may feel pressure to cut spend on a platform if the fact that they are advertising there could cause harm. Brand advertising is an exercise in optics, and those optics can extend not just to specific advertising placements but the commercial relationships that support them. That chain of perception matters. And it’s an acute risk for ad platforms that are primarily dependent on brand advertising revenue.

As I point out in Elon’s dilemma, Twitter is one such ad platform: the company revealed at an analyst event in 2020 that 85% of its advertising revenue was generated through brand advertising. The context of that disclosure is important: Twitter was making the case that it would reduce its dependency on brand advertising by building a direct response advertising platform. This aspiration is important because Twitter never really resolved it: as I point out in the piece, the platform substantially lags other scaled direct response channels, and that gap is likely to widen as the challenges of providing effective direct response tools become more complex as the privacy landscape changes and deterministic attribution is less feasible. Google’s PMax and Meta’s Advantage+ automation tools are perfect examples of these types of efforts.

I state in Elon’s dilemma that brand advertising revenue is akin to a participation trophy for a social media property; it’s a low-leverage means of converting attention into revenue. And while brand advertising revenue is fairly easy to harvest, it’s susceptible to disruption through advertiser boycotts and “shiny new object” syndrome. Direct response advertising tools aim to use behavioral and contextual data to unlock value through relevance targeting; brand advertising tools give advertisers untargeted access to attention. As a result, direct response advertisers buy conversions and revenue, while brand advertisers simply buy reach.

The ARPU metrics of various digital ad platforms make clear which form of advertising generates the most attractive unit economics. But this is also clear in looking to the evolution of these ARPU metrics over time. Meta’s ARPU growth rates in the US & Canada outpaced the global rate by nearly a factor of two following the introduction of App Event Optimization (AEO), an audience targeting tool that allows advertisers to bid against the completion of specific events (like purchases). This type of tool is wholly irrelevant to brand advertisers but is powerful in the hands of direct response advertisers, which aim to achieve direct and near-term commercial outcomes with their advertising spend.

The prospect of buying revenue is compelling: it invites the first dollar of spend from smaller advertisers that don’t have the budgets to establish brands through advertising. This is evident in the scale difference between Twitter’s advertiser base and Meta’s: Twitter served 2300 unique advertisers in August of this year and 2900 in September, whereas Meta has boasted that more than 10 million advertisers buy media from its platform.

The disparity in risk profiles between Twitter, a brand-oriented platform used by single-digit thousands of advertisers, and Meta, a direct response-oriented platform used by double-digit millions of advertisers, is apparent in the consequences faced by those companies from advertiser boycotts. In Meta’s Q2 2020 earnings call on July 30th, 2020, at the tail end of an advertiser boycott, Mark Zuckerberg said:

Some also seem to wrongly assume that our business is dependent on a few large advertisers. While we value every single one of the businesses that use our platforms, the biggest part of our business is
serving small businesses. Our advertising is one of the most effective tools that small businesses have to find customers, to grow their businesses, and to create jobs.

The impact of the advertiser boycott on Meta’s business was not material: the company’s advertising revenue increased by 10% on a year-over-year basis. The New York Times reported at the time:

Many of the companies that stayed away from Facebook said they planned to return, and many are mom-and-pop enterprises and individuals that depend on the platform for promotion…Of the 60 percent of DEG clients that joined the July boycott, four out of five are planning to return to Facebook in August, with many having “decided it’s too much for them during a difficult economic time to remain off,” Ms. Sheek said.

Contrast this with Musk’s statement about the impact of conscientious brand advertiser pullback. A scaled, diversified advertiser base is more insulated from any sort of coordinated action than a concentrated advertiser base, and brand-oriented ad platforms are designed to absorb the budgets of the largest advertisers in the world. Ultimately, this is a risk: brand budgets are fickle and, because they can’t be quantified directly to revenue, they represent a convenient path to cutting expenses in the case of an economic downturn.

Photo by Taras Chernus on Unsplash