Q&A: Reactor’s Audio Testing Capabilities

Since co-presenting—with Wavemaker Global—on Implicit testing of voices at this year’s Cannes Lions Festival of Creativity, people have asked a number of questions about testing audio implicitly. We’ve gathered some of the answers to the most frequently asked questions on implicit testing of sound.

Q: What kinds of audio are you able to test?

A: We can test almost any form of audio. This includes soundtracks (e.g. for TV ads, or web ads), voices, sound logos, sounds that products/apps/software make, radio ads, etc. Typically we test shorter formers of audio, so if you have audio that is longer than several minutes we would advise cutting it down to a representative selection.

Q: How does the process work?

A: Once the test is set up, most clients run it online. The minimal sample size for one of these tests is 200 people, so running online is a cost-effective way to recruit people, and they can take the test on either their home computer, smartphone or tablet device. Most tests last 10 to 20 minutes and usually involve the following steps:

  1. Participants are recruited online and typically answer a questionnaire that will qualify them for participation in the research.
  2. First checking that the person’s device can play audio and its speakers are working.
  3. Sometimes the context of the audio will be explained (i.e. along the lines of ‘You are about to hear some sounds that could be used in a smartphone app.’ or ‘You are about to hear a radio ad’), and this may include playing the full audio selection.
  4. Next we begin running the implicit testing exercises. These typically involve a timed, fast sorting task during which sections of audio will be played in a randomised order.
  5. If relevant, several survey/questionnaire questions can be added after the Implicit exercises.
  6. Once at least 200 people have completed the test, the data is available to download from the Reactor platform.
Q: In which countries are you able to run audio tests?

A: Reactor tests can run anywhere in the world where potential participants can get an internet connection. Of course, internet connection speeds and penetration of ownership of computers and mobile devices does vary around the world, but, at the minimum, if you can get enough people with smartphones and internet connections then a test is possible.

Q: What do you need in order to conduct a test?

A: The audio files and words that describe the feelings, emotions or qualities that you want your audio to evoke.

Q: What are the benefits of testing audio implicitly?

A: We believe there are several benefits to testing audio Implicitly.

First, our reactions to audio are often emotional and spontaneous. Not only do people differ in their ability to understand and report their own emotional reactions, but the mere act of thinking about them consciously in order to answer an explicit question, may change the nature of their reaction. In other words, just asking people to rate or describe their reactions to audio may not give an accurate reading on people’s true responses. This problem can be made worse by the fact that questionnaires asking people to rate their responses to audio testing can be somewhat boring for participants and many are tempted to just randomly speed their way through, something that is not possible in an Implicit Response test.

Second, many audio tests involve playing multiple sound options that may sound similar. When asking people to report their reactions to things that are very similar, we find that people are often unable to do so. This is perhaps because they rationalize that they are being presented with similar options and don’t wish to appear—even to themselves—to be swayed by apparently subjective and small differences.

Last, the outputs from our implicit audio tests are quantified, and hence comparable across different groups and also amenable to statistical analysis, meaning that you can measure the degree of confidence with which different audio options might be creating specific reactions.

Q: What other questions can you answer with implicit audio testing?

A: Most of our implicit audio tests involve testing the ability of audio clips to automatically evoke different attributes. These attributes are typically words that describe the feelings, emotions or qualities that you want the audio to evoke. However, another form of test we can conduct is measuring the degree to which a section of audio most naturally fits with a brand. If you have other types of research questions about audio, please feel free to talk to us about them and we may be able to design a new test to fit your needs.

“The tongue paints: What eye’s can’t see: the power of voice” was held on June 19th at 10.30am, at the Palais II stage, at the Cannes Lions International Festival of Creativity.

For more information, visit: