“We’re only focused on conversions.”

“Sales are our only KPI.”

“Attribution moves the needle.”

These are things I’ve heard over and over again talking to agencies and brands about their podcast and digital audio ad metrics. It’s usually during a conversation where they’re explaining why they’re not interested in brand lift. 

I get it. If you’re selling a product, sales pay the bill, and without revenue, you don’t have a business. I’m in business, too. I have a marketing budget, and I also want to measure the effectiveness of those dollars. Attribution provides crucial data connecting that ad impression to real-world consumer action. 

I’m also in the brand lift business, and I fully acknowledge that brand lift is not a substitute for or a challenger to attribution. 

Instead, brand lift is a powerful complement to attribution. Brand lift is chocolate to attribution’s peanut butter. If attribution is a sunny beach, brand lift is a cold Aperol Spritz in your hand 🍹.

The Game-Changer of Attribution

I was working in research at Midroll Media when podcast attribution first burst on the scene, and it was a game changer, especially for direct-to-consumer brands. Interestingly, we had already been doing brand lift studies for a couple of years, which helped open the door to major consumer brands. But the introduction of attribution helped to drive even more investment in the medium – I was in the center as it happened.

During my time at Midroll and Stitcher, and through our acquisition by SiriusXM, I oversaw attribution measurement for hundreds of campaigns. I maintained an enormous spreadsheet tracking every single one of them, so I could see the ones that topped the list, or sank to the bottom. Very often, I was asked by our sales team and agency partners to investigate and explain the performance they were seeing with a campaign. Full disclosure: I was asked to analyze poor or disappointing performers way more often than winners.

“Why aren’t we seeing enough web landings?”

“Why is this show driving fewer conversions than the other one?”

“Why is this host-read resulting in fewer sign-ups than the announcer-read?”

I was always glad to answer these questions, because it meant that the buyer and the advertisers were looking below the surface. Too often attribution results are used to cut anything that seemed to underperform. I heard from so many frustrated salespeople when a campaign was pulled based upon attribution results, when they knew the story wasn’t as simple as, “that podcast isn’t working.”

Explaining Poor Attribution Performance

When attribution showed poor performance, the first place I’d look is the ad itself. I’d get all the scripts and copy points, and the audio from all the ads, and then dig in. Sometimes the answer would be clear. For instance, one ad only mentioned the URL once, while another mentioned it three times. Not surprisingly, the latter drove more web landings. 

(Quick aside: I remember one time when the ad copy provided by the advertiser didn’t mention the URL at all. And they wondered why it underperformed for web landings.) 

But it wasn’t always so simple.

Other times I had to listen closely to the ads to understand what the listener actually experienced. Testing thousands of podcast ads, I developed a pretty good sense for what makes a good one. And where did that understanding come from? It came from analyzing brand lift results.

Listeners Will Tell Us Why Ads Do and Don’t Work … If We Ask

That’s because brand lift is the method where listeners tell us what they remember from an ad, and how it changed their perception of the brand and product. Malcolm Gladwell talks about the value of dedicating 10,000 hours to achieve mastery – what about analyzing 100,000 listener responses? That’s what I did (and still do), and it’s helped me identify the elements of an effective podcast ad – what listeners pay attention to, understand, remember and act on. 

That’s all well and good, but so many times I wished the advertiser had run a brand lift study side-by-side with attribution. While I like to think I’ve got a pretty well-tuned instrument, I don’t claim to be a psychic, and I’m not able to mentally model every listener’s and every audience’s response. I can present some pretty plausible hypotheses about why and how an ad worked, or didn’t, but a brand lift study running at the same time as attribution would provide concrete evidence.

Brand Lift Tells Us About the Consumer Before They Take Any Attributable Action

A brand lift study allows you to identify what happened before that first click, web landing, or purchase. Most listeners who hear an ad don’t immediately open their browser or wallet. And until they do those things, those listeners are invisible to attribution. But the ad is still influencing their action (or inaction). So, if they’re not visiting the website, signing up, adding to cart or checking out, you can hazard a guess why, but you won’t know.

Brand lift starts with awareness. It’s logical they won’t visit your website if they don’t know your  brand, and brand lift tells us that. What are the other prerequisites to  purchase? Brand favorability, consideration and the intent to purchase. Brand lift tells us that, too.

The best moments in campaign analysis come when a campaign has both attribution and brand lift data. We can see things like:

• Women liked the creative and men didn’t. That’s why the male-skewed shows underperformed on all metrics, including attributed conversions. It wasn’t a mismatch between the brand and podcast, it was a mismatch between the creative and the audience. 
• The campaign performed for brand awareness, but didn’t drive agreement with brand statements, or generate web landings. That’s because there were too many copy points, overwhelming the listener rather than educating them and leading them to purchase.
• The announcer-read outperformed the host-read in web landings. This was coupled with  an actual drop in brand favorability among those exposed to the host-read ad. In this case, the host wasn’t permitted the same kind of creative freedom they normally enjoy.

Without brand lift results informing the why behind the what, those advertisers probably would have cut the underperforming shows and placements from the campaign. And they would have sacrificed reach and incremental sales. That’s how brand lift measures, informs and optimizes true return on investment.

Ultimately, isn’t more performance data better than less? The Signal Hill Team has helped dozens of brands better understand their audio campaign performance, and optimize for better results across the full marketing funnel. Hit us up to find out how we can help you.