As part of the WARC x Spotify “Getting your brand heard” report, Norman Wagner, Head of Group Media at Deutsche Telekom, discussed what active versus passive attention tells brands about users and how Deutsche Telekom has successfully applied digital audio within its media mix.
Norman Wagner, Head of Group Media, Deutsche Telekom
What are the key goals or KPIs you’re looking to achieve through your advertising and marketing?
It depends on the campaign’s strategic goal. The goals are reach and frequency if it's a brand campaign or a product launch. Here, we need a mindset shift. We would look at conversion as a goal lower down in the funnel. The KPI at the end is always increasing sales. Overall, this also depends on channels. For instance, for video, the KPI is view-through rates and reach.
That has implications for the scale of your marketing and the level of targeting. How does this factor into your considerations and your planning?
Usually, we have wide audiences on telecommunication. We tended to limit this to age in the past, typically using socio-demographic targeting. We don’t have this problem in digital audio, for instance, with older audiences. Oftentimes, for our standard products, we do less and less targeting. But that changes when you have specific products, like the ones for younger audiences. We try more contextual targeting in these cases.
For audio, we also try to be in the right environment – less music and more show formats with spoken word, for instance. Podcasts can get very interesting because as a brand, you are in a completely different user situation.
What does your media selection process look like to achieve different goals?
You need a strategy about what you want to achieve, right at the beginning. Decide where you can have which impact and also what kind of message you want to put out. If it's just a short message like, “There's a new plan now, get the iPhone 13 for €1”, you know, it's just about frequency. However, when you want to tell a story, you need much longer formats. Then it completely depends on which channel can deliver that format and reach your goal.
When you're thinking about that relationship, how much testing or experimentation comes into play? Is there a separate budget for this?
We continuously conduct brand surveys and do brand tracking every week. We believe in continuously investing on brand tracking. Additionally, we run internal media mix modelling which also helps us identify media channels.
We do the strategy, media mix and media analytics in-house. This makes us independent in selecting which media mix we're going to use. The agency then optimises within that channel.
The telco market is an ever-evolving space in terms of products and services. How does your media selection evolve in line with the industry?
I think plannable effective reach is our biggest issue. There’s a loss of reach on media like linear TV. It’s also especially difficult to make reach and frequency planning possible. It’s also challenging and nearly impossible to have cross-media reach and frequency reporting. We are working to solve this.
One more big challenge in digital is that we are in an ecosystem of silos. So, how do we optimise reach across Spotify, linear radio stations and their web radio stations, etc.? This is a massive problem.
How do you think that the siloed ecosystem is affecting your thinking about a balance between brand and performance?
I think the digital ecosystem has been built around performance metrics. What you see currently is an ecosystem which can identify demands. For instance, say when someone does a Google search or visits a website about telecommunication, etc. What’s missing is the technology and the ecosystem which creates demand. It’s how you grow as a brand.
Let's say, on YouTube, for example, you have skippable ads. When someone doesn't have the demand or doesn’t like your brand, they simply skip your ad. It’s the same on audio. If you don't like it, you just don't listen to it or you change the track.
Not being able to create demands because the tools don’t exist is the digital ecosystem’s biggest problem. When you do a digital campaign and let the algorithm work on optimising reach, you reach as many people as possible but just once. This means that 80% of the reach is with just one contact; it burns money. The tools aren’t ready where we can have five to ten contacts. The digital mindset is all about frequency capping, but we would need a mindset for effective reach instead. The algorithms, too, help us reach as many as possible within the silo. And that's a problem.
How are you currently using audio within your media mix?
We had a very successful big campaign where we fought against hate on the internet. We spent two-digit millions on it. Usually, you have a film for this, but audio very works well.
Audio is capable of building reach quite fast. Hence, we use podcasts, linear audio, web radio and Spotify. We see this working well in our brand tracking. Audio is also very fast creating a sales demand, too. I do think the medium can work very well for brands; you just need to blend different strategies.
You mentioned audio’s strengths in terms of reach and brand. What are its weaknesses or areas of improvement?
The databases for audio or radio are pretty poor in several markets, especially when it's non-digital. You're usually on interviews which are done half-yearly.
I recently spoke to one of the big radio broadcasters, who said, “We know we could use digital data. There are new methodologies like audio matching on smartwatches, etc. We are testing that to see how close it is to the truth.” But I want to add, “You're not looking how close it gets to the truth, you're looking how close it gets to your old data.” Because what is the definition of the ‘truth’? That’s missing.
So, what's the real reach? You know it works when your campaign is big enough. However, you don't understand because of the market data but because of the data based on sales, brand lift and awareness.
Do you think there are any unique strengths of audio which are underused?
I think that the medium’s speed is very high. Audio doesn't need a long time to work; you have most users’ daily usage. You can get plenty of reach within a day when you do a big campaign.
The typical situations are in the morning, say when people go to the office in their cars. That’s primetime; when you invest money, you achieve big reach quickly.
There’s been some research around active attention versus passive attention. How has that factored into your thinking around audio or just generally, your media selection?
On podcasts, we have very active attention. With this, you can have more complex messages which are otherwise very difficult to get.
We also look at the time to identify the media usage situation. For instance, the audio usage of music is different during the day than when one is driving. We look at the session length to find out how long it has been running. If it’s on for hours, then probably it's been playing in the background.
I want to have targeting options on usage situations. For instance, when you're on Spotify or when you are actively searching. That’s a highly active user and a completely different one compared to one that hasn’t touched the app for two hours. But the gap in the targeting ecosystem is that you can’t have these differentiations.
How do you see the future for audio and how telco brands might adopt it?
There is still a problem with scalability on podcasts. They need a good reputation as an advertiser; the host of the podcasts can refuse to associate with you.
When you're a brand with a bad reputation, currently, for example an oil, tobacco or weapons company, there might be some problems.
For audio, we will go into digital audio more and more. This means we need measurement options across the silos and if that's not coming, I think it will become difficult.
