What Thriller teaches us about advertising audibility

What Thriller teaches us about advertising audibility
Swedien's logic was that if Thriller sounded good on basic speakers, it would sound good on every other device

What does Michael Jackson’s Thriller have to do with audio advertising?

The story goes that, if you play the song on any device, whether it’s a smart speaker, laptop or through headphones, it will sound equally good, because Bruce Swedien mixed the song via very basic speakers.

Swedien’s logic was that if the song sounded great on very basic speakers, then it would also sound great on every boombox or any other device, as he noticed most people at that time did not have advanced hi-fi systems at home.

This led Michal Marcinik, AdTonos’ CEO, to question why this approach was not used for audio advertising, especially when there has been such a proliferation of devices and audio types beyond radio.

In turn, this inspired research into how to enhance audio ads for different devices, the impact on ad effectiveness and a reappraisal of how the market defines “audibility”.

Audibility standards need improving

Marcinik argues that the standard of “audibility” in the audio advertising industry needs to be improved and there is “space for some industry-wide discussion”.

“We are challenging the Media Ratings Council (MRC) standard, which says that actually the ad is viewable or audible if it’s consumed by two seconds, no matter how long it is,” he told The Media Leader.

“There’s a space for improvement, because if you have a 30-second-long ad and a five- or six-second-long ad, why should you put the same measurements to both? It should be a percentage; it should be based on quality of listening or quality of audibility, instead of putting the fixed amount of time to every single creative you have.”

Marcinik added that the definition of a minimum level of volume for audibility also needs to be changed, since different devices have different ranges of volume. For instance, a TV has 100 volume levels; if it is on level one, which is technically defined as “audible” because it is above zero, it is in fact quieter than a ticking watch.

Audio has been overlooked

The MRC standards for viewability have been a topic of debate in online display, video and TV advertising in recent years.

For Marcinik, audio is “a little bit overlooked” in the conversation around MRC standards and also in attention metrics.

“We decided to bring proper attention to how the ad sounds, because it has a direct impact on the audibility and we prove that through our scientific research with MRI scans that there’s concrete evidence that, through enhancing the quality, we could actually extend the ROI,” he explained.

“If the ad is audible in a much better way, if the quality of the sound is at its highest, then audibility is also performing much better.”

AdTonos proposes adding two conditions to audibility: the ad must not be filtered or flagged as invalid traffic and the volume of the device through which the ad is delivered must exceed 10 decibels for 25% of the ad’s length.

An IAB and GroupM report, The Evolution of Digital Audio Advertising in Europe, found that 37% of the buy side and 43% of publishers surveyed cited audibility as “an important metric” to evaluate digital audio campaigns.

Like a blurry movie

An ad’s audio quality is “sometimes unfortunately overlooked” and not every ad is being produced with every device considered, stressed Marcinik.

He also highlighted that before programmatic audio, there was only radio advertising, so ads only needed to be produced in a way that sounded great on radio.

For Marcinik, not changing or enhancing the audio for different devices is akin to watching a blurry movie where you cannot see what is happening.

“We believe that the quality of the sound should be adapted accordingly to how those devices process the sound,” he explained.

“We are quite certain that it benefits all the players on the value chain, from DSPs [demand-side platforms], buyers and publishers, because where we are right now with the programmatic ads which are received, not every single ad might be the same good quality. So it’s for the benefit of all the people involved in digital audio advertising — especially, at the very long end, the listeners themselves.”

Neuroscience support

Every device processes sound in a different way based on technical limitations. A smart speaker works differently to, say, Apple AirPods earphones or a laptop.

AdTonos carried out research into the most popular audio devices on the market, what their volume levels are and how they process sounds, as well as how people listen. The top devices for audio listening were found to be Amazon Alexa, Samsung TVs and Apple’s MacBook Pro 13 and AirPods.

To establish how different devices processed sounds, researchers played a sound sweep of every single frequency audible to humans. This allowed “deficiencies” on each device to be spotted and then the frequencies could be adapted to make them more audible.

From this, AdTonos developed an Amplifier tool to ensure an ad sounds “equally good or best” across the most popular devices.

AdTonos then commissioned a neuroscience study with Neurensics to understand how human brains react to enhanced and non-enhanced audio and the effectiveness of the ads. It involved a blind test with randomly selected ads from its programmatic server and an MRI scanner.

Ads were “enhanced” for earbuds specifically and participants were played this version and the original “non-enhanced” version in an MRI scanner to detect how they registered based on general advertising benchmarks developed by Neurensics for positive and negative emotion, effectivity, likeability and annoyance.

Non-enhanced audio ads registered more negative emotions (-0.46), while enhanced audio triggered more positive emotions (0.56).

Left chart shows enhanced ads, while right chart shows non-enhanced ads. Source: AdTonos and Neurensics


In addition, enhanced audio ads had higher effectivity (0.42 for enhanced compared with -0.57 for non-enhanced), higher likeability (0.64 vs 0.47) and lower annoyance (-0.41 vs 0.44).

Marcinik said of the results: “The beauty of this is, actually, you cannot cheat that, because this is exactly how our brain subconsciously react. It’s how the human brain reacts to advertising.”

Adwanted UK are the audio experts operating at the centre of audio trading, distribution and analytic processing. Contact us for more information on J-ET, Audiotrack or our RAJAR data engine. To access our audio industry directory, visit audioscape.info and to find your new job in audio visit The Media Leader Jobs, a dedicated marketplace for media, advertising and adtech roles.

Media Jobs