Advertisers often lament the lack of transparency they get from automated advertising tools.
Although they’re wooed by the promise of performance, they’re wary of losing visibility into the workings of so-called black-box systems, like Google’s Performance Max and Meta’s Advantage+, both of which rely on AI and machine learning to help advertisers optimize their bids.
But advertisers shouldn’t be so obsessed with control just for the sake of having it or out of fear that a platform is holding back information, said Ben Yahalom, president of True Classic, a DTC casual menswear brand that now spends the majority of its Meta budget on Advantage+ Shopping Campaigns.
“I think everyone is a control freak to some extent,” Yahalom said. “But I like to go deeper and ask myself this: Is my desire for control going to help me with performance or not?”
‘Wherever performance goes, our budget flows’
True Classic is a dyed-in-the-wool performance marketer.
The brand is “solution-agnostic,” Yahalom said, in that it’s open to spending on any channel using any tool in any market, so long as it helps achieve True Classic’s three main objectives: acquiring customers, driving sales and encouraging repeat purchases at a break-even price point.
“If we can do that, we’re beyond happy and we’ll do it all day long,” he said, “and that’s because it means we’re maximizing a given channel so that it’s operating as efficiently as possible for us.”
True Classic applies this approach across the spectrum of its paid advertising mix, which spans Facebook, Instagram, Google Search, YouTube, TikTok, podcasts, TV, direct mail, email and SMS.
“Wherever performance goes, our budget flows,” Yahalom said. “We’re really not playing favorites here.”
Made to measure
But to feel confident that a channel or a solution is performing, True Classic has to be rigorous about measurement and attribution.
The company is working with and testing a portfolio of different measurement providers, including Triple Whale and Northbeam for multitouch attribution and Measured for incrementality studies. True Classic also collects data from post-purchase surveys and is currently onboarding a partner to help with media mix modeling, which involves using statistical techniques to analyze historical data and make predictions about future performance.
“We’re very numbers-oriented, so the approach we take is rooted in a deep analytical understanding of what we’re spending and exactly what we’re getting from it,” Yahalom said. “That means we can’t stop at CAC [customer acquisition cost] or ROI. We really need to look at the entire cost structure.”
True Classic analyzes everything from production costs and fulfillment fees to shipping, returns and exchanges because margins can change depending on the situation.
For example, without considering how much it costs to mail a True Classic T-shirt to someone and process an exchange, if necessary, it’s not possible to have a true sense of CAC. It ends up being more expensive to keep that customer happy than someone who didn’t attempt a return.
True Classic even coined its own metric to measure the impact that margin has on its return on ad spend, which it refers to as “CM ROAS.” (The “C” stands for “contribution” and the “M” stands for “margin.”)
Measuring return on ad spend is important, but calculating ROAS without accounting for profit margin doesn’t reflect reality.
“All brands really want to do is grow fast but also stay profitable,” Yahalom said, “and to do that, you need to consider your variable costs and have a clear picture of your unit economics; it’s surprising how many brands don’t understand their numbers.”
Man-machine
Which comes back to the question of control and the levers marketers should be pulling to manage their media in a way that drives performance.
Often, the sort of controls some advertisers yearn for, including digging through the data to find new optimization opportunities, are obsolete compared to AI-powered platform tools.
“You can’t outperform an AI,” Yahalom said. “It’s calculating, who knows, gazillions of parameters in real time, which a human can’t do.”
But there is a lot that humans can do to learn more about their customers.
True Classic, for example, conducts consumer research to examine the impact of specific ads, including demographic breakdowns for different segments and how they responded. It also collects post-purchase survey responses on its website.
All this information can be fed back into the platforms in different ways, including finding lookalike audiences – which sounds like a logical move. Why not try to reach more people who resemble your best customers?
The seemingly logical, however, is not always the most performance-focused and cost-effective thing to do.
For example, it might make sense on the surface for a brand to share a group of its highest-value customers with a platform to find lookalikes. But reaching a high-intent audience could turn out to be pricier – and less effective – than broad-based brand marketing, depending on the campaign.
“This is why it’s important for brands to ask themselves questions like, are they focusing on the right metric, the right setting, the right control,” Yahalom said. “There are so many fallacies in advertising, and some people miss what to do because they’re not willing to zoom out from a business perspective.”
