MarTech Interview with Paul Kelly, Chief Revenue Officer, A Million Ads

Hi Paul. Please tell us about your journey in the audio tech industry. What attracted you to this evolving space?

Great question. I don’t come from an ad tech background specifically, but it’s obvious that technology will continue to play a critical role in every and any advertising-related decision, from planning to verification to performance to creative. As a result, exposure to the ad tech ecosystem appealed, and audio appealed specifically because it has a unique set of dynamics at play.

Audio has a scaled incumbent business in the form of radio, so there are established buyers, currency, and proven performance vectors. But there are also new offshoot channels in the form of connected cars, speakers, and other technologies that open the door for new ad formats and products, like voice commerce and dynamic creative and new content formats like podcasts, audiobooks, and voice assistants.

In short: consumers have more opportunities to listen, and they’re wholeheartedly embracing it. We spend our lives with Airpods in our ears, or an Amazon Echo speaker playing in the background, or in the car on the way to work. Unlike video or digital, audio is now something we can experience passively and pervasively anywhere and everywhere.

This is a unique confluence of explosive innovation and growth on top of (versus at the expense of) an established channel and marketplace. That mix is what attracted me to audio tech.

Tell us more about the vision of A Million Ads

Fundamentally, A Million Ads wants to make audio more effective as an advertising channel. Making that vision a sustained reality means appreciating the position of all the main parties: creators and media owners, advertisers, and listeners.

Audio advertising that is environment aware, and thus more relevant, timely, and interesting in terms of message, leads to better outcomes for all of these parties:

• Advertisers see their ads performing better due to raised appeal and better retention of the message, which increases the value of the inventory for both advertisers and publishers
• Users get a better experience, thus leading to growth in audience and aggregate consumption — and this creates value for creators and media owners.

What are the core offerings delivered by A Million Ads? What services do you offer to brands and digital agencies?

A Million Ads powers dynamic creative optimization (DCO) for audio channels — summed up as personalization. Our platform instantly configures and serves an optimized ad each time our studio receives a request.

Bid stream data gives us a real-time view of that listener’s environment (time, location, content being consumed, the device etc.) alongside any profile characteristics like loyalty status or interest/affinity.  Creative decisioning is guided by a pre-determined set of rules that ensures each listener is served the best possible message based on that combination of audience and contextual data.

In addition to this decisioning and subsequent ad serving, we also script and produce the ads themselves for many of our clients including Target, American Express, Under Armour, Uber, and many more.

Much research is happening on AI and machine learning’s role in transforming branded audio content. Could you tell us about your experience with these technologies and how they have influenced audio marketing?

We have worked with a few synth voice AIs already — they’ve already achieved being indistinguishable from human voices. But the main use case from an advertising perspective is being able to mimic celebrity voices and thus create a number of versions that would otherwise be too time- or cost-intensive.

Combining generative AI like Google’s Bard with synthetic voice AI could absolutely prompt a flurry of experimentation in the branded podcast space, or even lead to the creation of branded radio!

How can companies plan their investments in audio for upper-funnel campaigns?

The most crucial step is being in the game. It’s easy to get lost in the minutiae of campaign planning and messaging and ultimately loses sight of the big picture. Namely, audio is a uniquely capable medium for upper-funnel campaigns.

There are a few reasons why that’s the case. Reach is a big one. Most people interact with audio by listening to a podcast or internet radio, tuning into commercial broadcast radio, or simply through a streaming service like Spotify or Apple Music. With so many formats, there’s this massive addressable market for advertisers to talk to.

But perhaps more importantly, audio offers intimacy and memorability that you don’t get through other mediums. Suppose you hear an ad read by your favorite podcast host, for example. In that case, you’re more likely to remember it than an ad where the narrator is a professional voice actor hired by a casting company. There’s a strong connection between the creator and consumer, and brands can leverage that.

That’s not to say that “non-native” audio ads aren’t effective. They are. Looking at podcast ads more broadly, we see that 67 percent of listeners remember the ads and products they hear mentioned. There’s a stickiness that’s hard to find elsewhere. And that “stickiness” — paired with the massive audience sizes and intimacy — is valuable for upper-funnel campaigns.

Marketing Technology News: MarTech Interview with Mark Cattini, CEO at Acoustic

We are curious about understanding ad automation and how these capabilities can deliver better brand safety and ad fraud protection.

Naturally, brands are susceptible to brand safety. They don’t want their name to appear on content that doesn’t align with their values or may offend their customers. Any automated system has to take this into account.

With nearly 450,000 active podcasts, listening to every episode is untenable. Fortunately, the industry has automated systems that can tackle the hard work. First, we can glean a lot about the nature of the show simply by looking at the metadata. Internet radio shows and podcasts are distributed through massive centralized hubs — like Apple Podcasts or Spotify — where they provide information about themselves to help listeners find them. This metadata includes a description, genre, and so on. We can learn a lot from that.

We also have the tools to look at each episode individually, using speech-to-text algorithms and machine learning to identify potentially “brand unsafe” topics and phrases. Companies like iHeartRadio and Audacy are currently using these technologies, and they’re startlingly effective.

Please elaborate on your audience building and contextual data management.

At A Million Ads, we want to build the tools that allow brands to launch highly-effective advertising campaigns at scale without harming the listener’s privacy or experience. We rely heavily on contextual data — like time and location — to do that.

It’s important to note that this contextual data can come from many different places — not just the metadata we receive when a listener downloads a podcast. It could, for example, originate from the advertiser itself. A great example is the coronavirus vaccine campaign mentioned above. In one campaign we did for Benadryl, we created over 800,000 variations of the same ad, personalized for location, pollen count, and even the type of content the consumer was listening to.

This approach gets results but doesn’t feel particularly intrusive, unlike many other forms of digital advertisements. Each ad is personalized but without the prying that’s commonplace elsewhere.

Which brands are leading the pack in terms of delivering exceptional audio advertising campaigns?

That’s a great (and tough!) question. Over the past few years, we’ve seen a surge in advertiser attention in the audio world. A lot of money and focus is being channeled into this space, so we’re seeing some incredible work.

From my own experience, I’d have to say Under Armor. Working with digital marketing agency Digitas, we ran a campaign promoting their new Flow Velociti Wind sneakers. We felt it was vital for Under Armor to understand its target customer’s rhythms and behaviors and to incorporate that into the ad creative by using readily-available contextual data.

It was a strategy that paid off. With many brands battling “ad fatigue,” we minimized its effects by having contextual, relevant, and frequently-rotating advertisements.

In total, we produced over 50,400 possible advertisements and were able to help Under Armor achieve its strategic goals while speaking to audiences on an individual level. According to a brand effectiveness study commissioned from Attest Research, there was an 81 percent purchase intent for the Under Armor Flow Velociti Wind sneaker across audio, compared to 75 percent on video.

Tell us more about your digital marketing campaigns? What kind of results do you generate from your various B2B initiatives?

Dynamic creative works differently depending on the advertiser. A campaign for a travel-centric company will look different from one for, say, a telecommunications provider or a retailer. And I think that’s why it’s such an effective tool.

Dynamic creative lets you create a bespoke solution for each challenge, taking into account things like the objectives of the advertiser, the listening habits of the audience, and, indeed, the demographics of consumers targeted. Since we start with the expectation that we’ll be creating thousands — if not tens or hundreds of thousands — of ads, we have the freedom to think about what kinds of contextual data will drive listener affinity and advertisement relevance; a great example would be the pollen count data used in our campaign for Benadryl.

What does your Marketing Technology stack look like? Could you tell us what tools you use for analytics, automation, and content management?

We use a lot of proprietary code in our marketing stack and some industry-standard applications and services.

How can organizations quickly upgrade their audio advertising game plan?

The audio advertising industry is increasingly mature, particularly the technology and delivery side. You don’t have to look far to find advertising networks or agencies to help you craft and deliver your message.

But if an organization blindly enters this space because they feel it’s a missing link in their marketing strategy, they’re making a big mistake. Don’t get me wrong — I think audio is a compelling medium for advertisers. If you’re going to do something, you might as well do it right. As the saying goes: Failing to prepare is preparing to fail.

Before getting into audio, organizations should think hard about the actual messaging and strategy. Who are your customers, and where are they likely to be found? What do you want to say, and how can you articulate that in a fresh, relevant way that’s open to personalization? How will your campaign align with the attributes that make audio an effective medium in the first place — like memorability, as well as the powerful bond between host and listener?

They must get this right because the audio will only become more prominent, especially given the macroeconomic challenges afoot. As Netflix’s most recent financial statements show, people are cutting back on subscriptions. As a result, they’ll naturally gravitate towards ad-supported entertainment mediums like podcasts, radio, or Spotify’s free tier.

This will be a silver lining for brands because customers are moving from ad-free walled gardens to arenas to spaces where advertisements are an inevitable (and expected) element of the broader experience. There’s an opportunity for companies to build brand awareness and drive sales. Organizations have to get their strategies sorted — but they also need to move with agility if they want to take advantage of this moment.

What are the biggest challenges that could impact how organizations hire and retain tech talent? What do contemporary talent war trends look like to you?

We’re at a strange moment in time. We’ve seen massive layoffs at some of the world’s most recognizable tech companies, like Microsoft and Google. However, hiring continues, even at a slightly slower pace than before. It’s unclear whether the tech job market is in a state of boom or bust.

US tech job openings were up 25 percent in 2022. The layoffs witnessed at the start of the year are a rounding error in comparison. Hiring will continue to be a competitive (and expensive) game for the foreseeable future, but that could change in the months ahead if we see more large-scale layoffs like January’s.

Marketing Technology News: The Role Automation Plays in the Future of Ad Monetization