����AV

Among other things, Google’s search engine is the most successful billboard ever invented. Keyword marketing—i.e., placing text-based ads at the top of search results—has become an over $100 billion business responsible for the majority of Alphabet’s revenue.

Search marketing offers a variety of avenues for matching supply to demand. For example, companies can bid on their own keywords (e.g. brand and product names), putting ads in front of already-curious customers. Marketers can also bid on their competitors’ search terms, thus co-opting the brand recognition of others. The latter practice, known as competitive poaching, is increasingly prevalent, as casual Google users may have already noticed.

There are no hard and fast rules for winning the keyword wars. For researchers, it’s a tantalizingly uncharted area. Siddharth Bhattacharya, a professor of information systems at Mason, recently co-conducted (with Jing Gong and Sunil Wattal) . In a real-life experiment, the researchers tried out differently worded ads piggybacking on various competitor keywords, measuring performance based on click and conversion rates.

Across two distinct sectors—business academia and automotive sales—they noted similar results that help provide actionable guidelines for poachers and poachees. From the user perspective, their findings also contribute to understanding whether competitive poaching improves or detracts from overall search quality.

There were four varieties of ads used in the experiment, listed here in no particular order. First, ad copy stressing vertical, or quality-based, differentiation—expressed in phrases like “Top Ranked School. World Class Faculty” and “Best-Selling Luxury Car.” Second, horizontal-focused ads suggesting unique features and providing multiple options to consumers—e.g. “Extreme Versatility. Great Driver Assistance” and “Flexible Schedule. Mobile-Friendly Format.” Third, prescriptive copy conveying an aspirational appeal—e.g. “Have the Power. Push the Limits” and “Discover Opportunities. Leave Transformed.” Finally, control ad copy making no particular claim and serving as our baseline for comparison (“Upgrade to an Audi Today”).

No one type of ad performed best across the board. Rather, the effectiveness of the ads was directly related to the quality of the brands being poached upon, the researchers found. When poaching from high-quality competitors such as top-ranked business schools or luxury auto brands, vertical ad copy was up to 114 percent more effective than the control ad copy in driving clicks. On the other hand, for low-quality competitors, horizontal ad copy was up to 84 percent more effective in comparison to the control ad copy. Prescriptive ad copies did not show any effectiveness in driving click-throughs.

Bhattacharya explains the results by pointing out that different customer segments are after different things. Luxury consumers are sensitive to quality-based indicators, hence they respond to vertical differentiation. However, mass-market buyers have resigned themselves to not being able to afford the absolute best, but are looking for realistic options that meet their specific needs.

The prescriptive ads did not particularly resonate with either segment because their aspirational claims could not be verified. But that doesn’t mean prescriptive copy should never be used. For Bhattacharya, it comes down to the distinction between search-based and experience-based products. Search-based goods, such as the graduate degree programs and automobiles in the present experiments, are subject to objective evaluation based on their attributes. Experiences aren’t. A suspenseful, violent novel or movie, for example, may be well-made but nevertheless unappealing to those who dislike the thriller genre. More research is needed looking at experience-based products specifically.

The very first step when planning a competitive poaching campaign, Bhattacharya says, is to know whether what you’re selling is search-based or experience-based. Experiences are outside the purview of this research study. But for search-based products, your best bet is to tailor your ad copy depending on the quality of the poached brand.

The other determining factor of poaching success was physical distance. As you might expect, it proved difficult to poach from far away. Targets in close proximity, such as competing auto dealerships in the same town or city, worked better for driving clicks on poacher ads.

The experiment also investigated the presence of competition, i.e.ß, presence of competitor’s own ad appearing alongside the poacher’s ad on the search results page. Surprisingly, the presence of both ads increased clicks to the poacher—but only for high-quality keywords. Low-quality keywords resulted in fewer clicks on the poacher ad when the poachee was also present.

This implies that competitive poaching has mixed implications for search quality. On the one hand, it allows discerning high-end consumers to make more educated choices and comparisons between competing products. On the other, it seems to dissuade mass-market consumers, who may find the dueling messages distracting or confusing.

Bhattacharya believes the relevance of the study may extend beyond product marketing in the most strictly defined sense. Political campaigns are also avid keyword marketers. Vertical and horizontal approaches to advertising roughly parallel the ever-widening rhetorical divide between managerial elitism and working-class populism in many locales. If the analogy holds true, Bhattacharya’s study may offer a glimpse into the future of political messaging.