The game theory of branded search advertising

Branded search advertising is the practice of bidding for the ad impressions that are served alongside search results for keywords anchored to a brand (for instance, the brand’s name). It’s a controversial topic amongst digital advertisers: many, if not most, view it as a sort of “tax” that they must pay in order to be present on the platform that offers the search functionality, especially on retail platforms where users are presumably searching for very specific products to purchase.

A search for a keyword like “Pampers” on Google Search could be unrelated to commercial intent — the searcher may simply want to read about the history of the Pampers product or research its commercial ownership. That same search on an e-commerce platform is almost certainly motivated by near-term hygienic security. Allowing diaper producers other than Procter & Gamble to bid on the inventory run alongside searches for that keyword is difficult to defend as anything other than rent-seeking.

Branded search effectively weaponizes brand equity: it renders the investment made into brand recognition and awareness available to be co-opted at the point of a specific search. Companies aspire to build brands; branded search advertising holds that aspiration hostage. It’s difficult to not take a jaundiced view of branded search. It puts companies in the difficult position of having to defensively guard their brands — which they likely established assiduously, over long timelines, and at great expense — against capture.

A retail platform or digital storefront might defend the practice by declaring that its customers are consumers, not the retailers it serves, and that no retailer is entitled to wholly free access to its aggregated demand. A search for a specific keyword is an obvious signal that the consumer has a brand in mind, but allowing competitive ads to be served against search results doesn’t prevent the consumer from finding those products; it merely introduces more optionality to the purchasing process. And anyway, the platform should have no compunction against monetizing the demand it has accumulated and made available to its merchant clients. The consumer’s whims will ultimately dictate the platform’s product direction, and if it can monetize its search functionality without alienating the consumer, why shouldn’t it?

Both of these arguments are reasonable, but somewhat beside the point. In the same way that branded search influences all competitors in a specific vertical, so too does it influence all retail platforms in a specific commercial space. If one merchant bids against another’s branded keywords, then all merchants must, as a matter of competitive equilibrium; if one retail platform offers branded search, then all must, by the same logic.

The competitive reality for brand search is that it can be effective for advertisers in terms of incremental revenue contribution for their own keywords. In a fascinating paper on the topic titled Competitive Advertising on Brand Search: Traffic Stealing and Click Quality, Simonov and Hill find that, “Offense Is Effective, but Clicks Are Low Quality.” In the paper, the authors determine that:

The nominal metric, CPC [cost-per-click], suggests that focal brands enjoy a much lower CPC than competitors on their own traffic—in the case of one to four competitors, focal brands pay $0.52–$0.66 per click, whereas competitors pay $0.96–$1.56 per click.

The chart below depicts the same, with “Cost Per Incremental Successful Click,” or CPISC, representing the case when a consumer clicks a link and does not quickly (within 30 seconds) return to the search engine (in the case of this paper, Bing). Per the chart below, the paper finds that, “it is as expensive for a competitor to steal a click from the focal brand as it is for the focal brand to defend its traffic.”

In alignment with my argument above, the paper determines that the primary beneficiaries of the practice of branded search advertising are the search platforms:

In equilibrium, with competitors advertising and focal brands defending, the end outcome is that much of focal brands’ formerly free traffic is now navigating through paid channels, generating revenue for search engines every time a searcher clicks on the focal brand’s link. Although competitive advertising on brand search has been deemed legal—under the argument that such advertising increases competition and informs the end consumer—the results of allowing competitive advertising are a transfer of money from advertisers to platforms. Whether this transfer benefits or hurts the end customer depends on whether these extra costs are passed through to the end customer by advertisers and how the end customer benefits from seeing competitors’ paid links.

As unfortunate as advertisers might find the practice of branded search, their competitors are at a disadvantage for those keywords, and on a defensive basis, it’s possible to create incremental value. So how should an advertiser approach the decision? The simple chart above — which assumes that incremental value is possible to achieve through defensive branded keyword campaigns, which is my own empirical experience — attempts to map this logic. But there’s an important point to flag: the appetite for any given branded keyword is likely to be multitudinous, across many competitive players within the category, or even adjacent categories. So an advertiser with a branded keyword can’t just consider that their most direct competitor may bid on their branded keyword, but that many competitors will bid on their branded keyword, whether or not they themselves bid. This surfaces two important considerations:

  • The clearing price of the inventory may not increase at all as a result of participation by the brand owner, given the improved conversion rates for the brand’s ads against its owned keyword, and the fact that competition exists with or without them;
  • Assuming the brand owner and the competitors all bid at prices that are profitable / rational to them, given the conversion dynamics, the brand owner should win its branded impressions more frequently than its competitors.

The result of this: the brand owner gains more by buying its own branded keywords than it does by abstaining. The competition for its branded keyword is not constrained to a single competitor; that set of competitors likely has to bid higher to win those impressions than the brand owner does, given structurally lower conversion rates (because they do not sell what the consumer seeks). So the brand owner should be able to win impressions more frequently than its competitors, and if the revenue it generates from doing so is incrementally profitable, then it is rational to buy it. In other words: the brand owner’s benefit from defending its keyword is not only that it displaces its competitors from the inventory, but also that it generates incremental profit.

Branded search revenue is incredibly attractive to retail platforms, as I’ve articulated in my series that explores my Everything is an Ad Network thesis: because these products monetize existing assets (data), they essentially print free money. And although advertisers wish they didn’t have to defend their own brands, it’s likely inevitable that they’ll have to do so on every important digital storefront on which they operate. Precisely because brands intrinsically have value — not just to their owners, but to anyone obliquely associated with that brand’s category — then multi-party competition for those branded keywords will always exist, whether or not the owner participates in branded search.

Acknowledging this reality helps to inform the logic of purchasing branded search keywords defensively. It may be true that digital storefronts benefit most from branded search, but so long as the traffic acquired is incrementally profitable, brand owners do, too.

Photo by Jose Castillo on Unsplash