From smart speakers to interactions with technology on our mobile devices and computers, it obvious that a global ‘voice revolution’ is happening. What does that mean for market research?
Only good things. Voice makes it easy for researchers to have the best of both worlds. By adding voice to surveys, quantitative researchers tap into the qualitative world while providing respondents with a better experience. And qualitative researchers are able to quantify results at scale.
And, there’s more. Advances in voice tech, AI, machine learning, and computing methods have made it possible to leverage established behavioral science and accurately detect people’s emotions in their speaking voice based on their pitch, tone, and other characteristics.
In addition to uncovering people’s nonconscious emotional responses, because people give longer answers when they talk than when they type, especially on mobile devices, voice enables researchers to collect more, and richer, data. A recent A/B test revealed that voice responses were typically 4.5x longer than typed responses.
Researchers know the important role the nonconscious plays in decision-making. The ‘Holy Grail’ of understanding System 1 response is to detect them in real-time in a way that is reliable, authentic, and free of potential contamination related to the response being observed. Because voice can be a seamless add-on to survey platforms, unlike video analysis, it collects nonconscious data unobtrusively and therefore is less likely to introduce bias.
There’s an additional advantage to using voice in research. Because emotion is detected on the nonconscious characteristics of the speaker’s voice, when people ‘put on a brave face,’ as they might when talking about topics that make them anxious, the underlying emotion is uncovered. That’s different than sentiment analysis, which is based solely on the speaker’s words.
The rest of this post will focus on voice in quantitative research, where it is typically used for open-ended questions. Because getting more voice data is better for analysis, and because people who respond via voice are representative of the population, here are some practical tips on how to get more value from using voice in quantitative research:
1. Increase respondent engagement by enabling voice responses whenever you can, even if you don’t plan to do extensive analysis on those responses.
Across dozens of projects, we’ve found that, when given the choice of talking or typing, typically 40% to 60% of respondents will choose to talk. And, when people respond via voice they give longer responses—studies have shown anywhere from 35% longer to the 450% longer mentioned above.
2. Use voice on topics that are likely to generate meaningful emotion to realize higher value.
People don’t care deeply about everything. You’re more likely to receive value from voice on topics like health or politics than you would on a comparison of doorknobs or light switches.
3. Consider the audience, where people are likely to take the survey, and the sensitivity of the subject matter.
In general, audiences that are inclined to use a mobile phone are more likely to respond via voice. And, because people tend to have a more intimate/personal ‘relationship’ with their phone (think selfies) than with their computer, they often are more willing to talk.
But, if the subject matter is sensitive, people may limit what they say if they are being asked about a private matter in a public place. If your survey platform supports it, you might remind people that they can answer questions any time and come back to the OEs in a private moment.
And, it’s important to consider environmental factors at the location such as noisiness or distractions.
4. Encourage the use of voice.
From the first moment you introduce the possibility of voice responses your words make a difference. Consider, for example, the vast difference between, ‘We want to hear what you have to say’ and ‘Type your responses the usual way or speak your answer.’ You can introduce the use of voice as a benefit to the respondent because speaking responses is easier and 4-6 times faster than typing them.
Assure people that their voice responses will be treated with care (they are often protected by privacy statutes) but be careful to not go overboard with that as over-achieving on assurances can generate concern.
Take care with the way you present the instructions for using voice. Make them visible, simple, and clear.
And, consider making voice responses the default mode for responding to your open-ended questions.
5. Adopt a qualitative mindset as you phrase your questions – encourage storytelling and longer answers.
Voice enables you to inject qualitative research into a quantitative survey frame. Ask yourself, ‘How would I ask about this topic in a qualitative setting?’
Give people license to provide longer answers by using phrases like, ‘Respond as if you were telling a friend,’ ‘Talk for at least a minute,’ ‘Don’t edit your thoughts,’ ‘Don’t stop at one thing,’ ‘Give an example or two,’ or ‘Look at different angles.’ For certain surveys/samples, you might even consider something like, ‘Speak until you’re exhausted.’
6. Use warm language.
Consider the difference for the likely response between something brisk like, ‘Why did you pick that number?’ and ‘We’d like to hear more about what you think about the brand’ or ‘Imagine what could lead to a better score and tell us more.’
7. Shift the response mode from typing to voice.
At a minimum, have the software read the questions to the respondents. If it makes sense in your survey design, consider ‘queuing’ your open-ended questions and encourage respondents to switch to speaking to answer them. Also, consider placing follow-ups at the end of the survey so people can share more of their opinions.
8. Choose voice software that supports your full project cycle.
Look for software that integrates with your preferred survey platform(s), quantifies emotion, makes it easy for your team to quickly explore responses, and supports your client presentation well. Also look for simple comparisons of positive and negative responses, detail by type of emotion, and scores that reveal intensity.
If your work includes multiple languages, look for software that supports them. Consider getting transcription, sentiment, and emotion from a single supplier to improve speed-to-result, enable comparisons, and lower cost. If transcription is included and your work is an industry with a specific vocabulary or if brand names are included (e.g., if you are surveying car enthusiasts the software should hear ‘Guilia’, not ‘Julia’), look for a supplier who supports that vocabulary.
Nonconscious analysis of people’s emotions via voice has the potential to add a valuable new dimension to your quantitative – and qualitative – research. Its application to research is new and best practices are evolving. To maximize the value you receive incorporate tips like those above and work closely with your voice supplier to tap into their expertise.
The post 8 Ways to Maximize the Value of Voice in Your Research first appeared on GreenBook.
source: https://www.greenbook.org/mr/market-research-methodology/8-ways-to-maximize-the-value-of-voice-in-your-research/?utm_source=rss&utm_medium=rss&utm_campaign=8-ways-to-maximize-the-value-of-voice-in-your-research
Your content is great. However, if any of the content contained herein violates any rights of yours, including those of copyright, please contact us immediately by e-mail at media[@]kissrpr.com.