By Prassath Leelakrishnan, VP of Consumer, Ethos
Imagine casting a fishing net in the endless ocean, hoping to snag a few prized marlin. That’s the familiar yet frustrating challenge faced by marketers on performance marketing platforms trying to find buyers for their high ARPU products. Platforms like Google, Meta, and Tik-tok serve as the preferred fishing grounds for performance marketers. These platforms excel at attracting vast schools of fish through their algorithms, designed to match ads with potential buyers. They are very effective for mass market and high frequency products like sneakers and gaming apps because there are hundreds of millions of prospective buyers and the platforms have numerous signals to find the most likely buyers. On the other hand, buyers for high ARPU products like SaaS and financial services are like Marlins. They are fewer and ad platforms also don’t always have high fidelity signals to spot them because of a lack of past purchase data for such products. Therefore, the efficiency suffers, resulting in low Return-On-Ad-Spend (ROAS). These advertisers would need to customize the off-the-shelf fishing gear offered by ad platforms and deploy tailored strategies to catch their marlins efficiently.
More bait to feed the algorithms: Milestone based signals
Ad platforms rely heavily on the purchase signals sent back by advertisers. Their machine learning algorithms use these signals to determine similar users that are likely to purchase the advertisers’ offering. These purchase signals are like chum attracting the fish. Now, consider two companies, each with a daily revenue of $100,000: one markets a $100 sneaker, the other offers a Wealth Management service worth $10,000 in LifeTime Value (LTV). Despite each generating $100,000 daily, their dynamics vary dramatically. The sneaker brand would sell 1,000 units daily and transmit 1,000 signals to the advertising platform. The Wealth Management firm would just emit 10 purchase signals a day from signing up 10 customers to generate the same $100,000 in LTV. This stark 100-fold difference in optimization signals is akin to having 100 times less chum to attract the fish.
The good news is that high ARPU companies can artificially manufacture the chum that can help attract the big fish. While companies with high ARPU products often have low purchase volumes, they usually experience significant website traffic and progression through the purchase funnel. As users advance to distinct stages of the purchase funnel, each milestone reached signifies a deepening intent to purchase. Companies can classify key milestones as events and share with ad platforms as purchase signals. These milestone based signals, while less precise than final purchase signals, increase the volume of actionable data shared with ad platforms and enhance the platform’s ability to identify other potential high-intent customers.
Casting the net in the right place: Lookalike Audiences
Imagine having a sonar that identifies and distinguishes schools of marlin. Lookalike audience targeting works similarly by enabling ad platforms to identify users on their platform that are similar to an advertisers’ existing customers. This is akin to casting the net in the right waters. Lookalike audiences are more critical for high ARPU advertisers because they help ad platforms narrow down target audiences, despite a shortage of purchase signals.
Lookalike targeting works out-of-box for most advertisers as they simply need to upload the list of existing customers and let ad platforms do their magic. A key limitation for high ARPU products is that there is a very high variability in customer value. For example, a customer investing $1M in a wealth management service is 10X more valuable than a customer investing $100K. Advertisers can account for this difference in customer value by establishing different customer segments and creating custom Lookalike audiences for each segment. They can do this by leveraging machine learning and statistical techniques that consider various characteristics such as LTV, website behavioral data, and demographics to create multiple homogeneous segments. Such a focussed targeting strategy will enable advertisers to attract the most relevant customers and tailor their bids for each segment to maximize their return on ad spend.
Perfecting your Lure: Rapid Creative Experimentation
Creative content has long been one of the most powerful levers to drive business results through paid marketing. Continuous A/B testing of creative enables advertisers to refine their message, visuals, and calls to action, ensuring their campaigns stay fresh, capture user attention, and ultimately drive superior results. The effectiveness of these tests is traditionally assessed by tracking variations in purchase volume or the ROAS attributed to each creative version. While this works for advertisers with high purchase volumes, the traditional approach of creative testing becomes impractical due to lower transaction frequencies. Using our previous example, the wealth management company with 100 times fewer purchases would need 100 times longer to test a new creative. Therefore, such advertisers often rely on top-of-funnel metrics like click-through-rates (CTR) to determine creative winners. The downside is that Top of funnel metrics like CTR can favor a click-baity creative over a creative that can attract high intent users.
Milestone based signals outlined above offer a strong middle ground between the low volume of purchases and noisy top-of-funnel signals. They showcase genuine user interest, while also providing sufficient volume of data to make rapid creative testing decisions and continuously refine the creative. This can enable advertisers to perfect their lure to attract their Marlins.
Fishing in diverse grounds: Multi-channel strategies
Relying solely on one fishing spot, no matter how rich it seems, limits your catch of Marlins. Similarly, for advertisers of high-ARPU products, where the target audience is smaller, focusing on a single marketing channel restricts the reach and potential customer pool. This is in contrast with mass market advertisers who can build large businesses by relying on a single platform like Instagram or Tik-tok for customer acquisition.
When high ARPU advertisers push to acquire more and more customers on the same channel, their efficiency decreases rapidly because the platforms need to cast the nets wider to find more customers. As customer attention is spread across channels, the cost of acquiring the next incremental customer may often be much lower on the next channel vs. going too deep on the first channel. As advertisers perfect their signals, targeting, and creative, they can often deploy the same strategies on those additional channels and achieve higher marginal return on spend.
The strategies outlined herein do not promise a big catch off every cast, but offer a framework for navigating the performance marketing seas. Success lies in the willingness of advertisers to experiment and iterate across the four critical dimensions: optimization signals, targeting, creative content, and channel selection. By continuously refining these strategies, advertisers can cast their nets with precision and improve their efficiency in reeling in their high value Marlins.
About the Author
Prassath is a growth leader with extensive experience in advertising and consumer businesses. He leads the direct-to-consumer division at Ethos. Prior to Ethos, Prassath built and scaled multiple advertising product lines at Meta and Twitch, where he helped advertisers improve their Reach and ROAS and created highly engaging advertising experiences for millions of consumers. He adopts a highly data-driven and experimentation-oriented approach to performance marketing.
