This week, Parag Agrawal, who just lately succeeded Jack Dorsey to helm Twitter as CEO, revealed a thread explaining the corporate’s strategy to combating spam on the platform. That spam, primarily perpetrated by “bots,” or representations of Twitter customers managed by software program, had been cited as a degree of concern by Elon Musk, who has agreed to purchase the corporate. Musk replied to Agrawal’s thread in a characteristically colourful vogue, questioning how advertisers might “know what they’re getting for his or her cash” if the prevalence of bots can’t be quantified with absolute certainty.
It’s essential in contemplating this query to know the composition of Twitter’s promoting income as a result of not each advertiser has the identical motivation or targets in buying advert impressions. I’ve written extensively (and generally, dismissively) about model promoting as a observe, however finally, there are two promoting ways that may co-exist in a type of yin-and-yang symbiosis beneath the broader umbrella of efficiency advertising and marketing:
- Model promoting, which seeks to make use of promoting impressions to construct consciousness for a product that can end in incremental purchases over some enduring, indefinite timeline;
- Direct response promoting, which seeks to make use of promoting impressions to generate attributable, incremental conversions (typically: purchases) inside some particular period of time from publicity to an advert.
As I suggest in What’s efficiency advertising and marketing?, model promoting just isn’t antithetical to efficiency advertising and marketing: model promoting and direct response promoting may each be utilized inside a efficiency advertising and marketing framework concurrently, or individually at totally different factors in a product’s or an organization’s lifecycle. The effectiveness of name promoting is usually tougher to measure than direct response, and generally that opacity of effectivity is used as a crutch to keep away from accountability by model advertising and marketing groups. “Constructing a model” is an imprecise and nebulous endeavor, and unimaginable sources could be invested into it. Model promoting could be pursued inside a efficiency mannequin, however fairly often it isn’t.
At an analyst day in 2020, Twitter revealed that 85% of its promoting income is generated from model promoting campaigns. This, by itself, just isn’t a trigger for concern or essentially an objectively unfavorable income construction. However the context of this announcement was that Twitter was wholly dedicated to altering that construction by bettering its promoting platform such that it might handle extra money from direct response promoting campaigns. Twitter made the purpose that it felt that direct response promoting represented a chance for the corporate: the economics of direct response promoting index to the worth generated for advertisers, and thus direct response income can scale in a manner that’s untethered to DAU development. Model promoting is just a operate of the scale of a platform’s person base: with model promoting campaigns, advertisers pay for attain on a CPM foundation.
Elon Musk is appropriate: bots forestall model advertisers from quantifying in a dependable manner the variety of individuals — actual, respiratory people — that see its adverts. However Twitter has a much bigger, extra urgent downside than bots. Twitter isn’t in a position to handle direct response promoting campaigns at an considerable scale. Fixing this, and welcoming extra direct response promoting spend to the platform, is a much more compelling income alternative for the corporate.
As a result of for probably the most half, direct response advertisers don’t care about bots. Direct response advertisers optimize in direction of enterprise targets: purchases, registrations, and so on. Bots could muddy a direct response advertiser’s understanding of its CPM metric, however CPM (and attain) is primarily an arrogance metric for direct response advertisers. Direct response advertisers are centered on “bottom-of-the-funnel” metrics that aren’t influenced by bots. Because of this, whereas annoying and unacceptable, the constant drumbeat of revelations that this-or-that massive advert platform inflated its attain metrics doesn’t are inclined to perturb direct response promoting budgets. I wrote an op-ed on this matter in 2019 that was not warmly acquired.
Final December, when Jack Dorsey stepped down from the CEO function at Twitter, I wrote a chunk titled Can Twitter turn into a direct response promoting behemoth? by which I questioned whether or not Twitter possessed the constitutional inclination and technical capability to develop its diminutive direct response promoting enterprise. From the piece:
Probably the most urgent query for Twitter’s new CEO, to my thoughts, is whether or not the corporate can create a direct response promoting behemoth that may rival different giant platforms like Fb and Google…Twitter’s O&O promoting platform is at the moment one thing of a chaotic, anachronistic system, and the corporate bought the advert tech asset, MoPub, that greatest captured direct response advert spend on cell. If Twitter sees promoting income, however particularly cell promoting income, as an integral part of the corporate’s development, then it ought to commit complete groups to promoting attribution and measurement. Nothing else ought to actually matter to the corporate: rising direct response advertising and marketing income needs to be Twitter’s first and most crucial aspiration.
Why hasn’t Twitter carried out this? There could also be structural causes: the Twitter feed is maybe extra frenetic and never as pleasant to direct response adverts as different product floor areas. But when the format isn’t the impediment to constructing an essential direct response enterprise, and really the hindrance is cultural, then addressing that’s virtually definitely a extra profitable initiative than completely eradicating bots, which is probably not potential and appears to be one thing of a white whale for Musk.