January 20, 2021
Eyes4Research
Humans are complex. In research, participants often find themselves limited to numbers and words to express themselves. Emotion AI goes beyond traditional research processes, interpreting human emotions for a more in-depth understanding of the topic at hand. This relatively new technology is taking the market research industry by storm, allowing MRX professionals to capture and analyze emotions for more informed decision making. Here is a glimpse into the three common uses of emotion AI and the benefits of this groundbreaking technology.
Text
Extracting emotional insights from text requires sentiment analysis- a process that uses natural language processing to identify and quantify positive/negative emotions from text samples. Natural language processing (NLP) harnesses machine learning algorithms and techniques to help researchers determine intent and categorize users. Sentiment analysis uses three main methods; rule-based, statistical, and hybrid. Rule-based sentiment analysis determines the emotions in a sentence by classifying text based on affected words. These words often correlate directly with a specific sentiment such as happy, sad, angry, etc. When using the statistical method, researchers uncover emotions using an ML model. This model helps identify the product and holder in a complex sentence. Finally, hybrid model sentiment analysis combines rule-based and statistical methods to detect subtly expressed sentiments.
Applications of this process include analyzing customer reviews, surveys, social media content, and other online materials. Using sentiment analysis tools can help businesses analyze KPIs and customer satisfaction levels, compounding and analyzing large amounts of online data to provide insight into overall brand sentiment.
Benefits:
Sifting through online data remains time-consuming and may not account for every piece of information on the web. Using AI technology saves researchers time finding and analyzing relevant consumer insights, helping unmask consumer sentiments and create data-driven decisions in less time.
2. Improve Social Listening
With a large amount of information available on the web, keeping track of consumer insights pertaining to a specific brand can be difficult. Utilizing AI technology simplifies this process, allowing researchers to more easily attain and analyze customer feedback to improve overall experiences and brand strategies.
3. Become More Agile
Staying up to date on customer behaviors and sentiments can help businesses identify trends and adjust quickly to changing preferences to stay ahead of competitors and provide exceptional customer experiences.
Video
Researchers can utilize AI technology through video platforms to monitor human emotions in relation to various stimuli. Facial movement analysis remains a frequently used measuring tool for identifying facial micro-expressions to uncover consumer sentiment. These tools work by identifying key features of the face to analyze changes in these features, attributing facial expressions to emotions. Through deep learning, AI video technology can recognize universal emotions, helping researchers go beyond traditional qualitative and quantitative practices. With this, researchers can analyze consumer engagement, emotional activation, and the overall impact of specific topics. Video AI technology uncovers insights in advertising campaigns alongside user experience, brand, concept, and product testing. Researchers can utilize this method in qualitative group settings such as focus groups or individual settings using personal computers and smartphones.
Benefits:
Unlike other forms of emotion AI, video analysis allows researchers to track eye movements, identifying areas of focus to gauge which elements draw in research participants. This becomes very useful when testing advertising campaigns, as this data provides insights into the effectiveness of specific design elements.
2. Versatile
As mentioned above, researchers can use video to measure emotions in both group and individual studies. With this, video can be used alongside methods, such as audio, to capture emotions from both verbal and non-verbal indicators.
Audio
Researchers have begun using AI technology to decipher emotion in audio. Speech emotion recognition (SER) processes and classifies speech signals to detect and analyze the sentiments of a speaker. These programs go beyond words spoken, analyzing vocal tone to determine emotions.
Benefits:
When conducting qualitative studies, emotion AI technology helps researchers dive deeper into the responses of participants. Analyzing emotions through vocal tones concerning specific topics, words, and phrases provides a better understanding of consumer intent and emotion for more informed and impactful decisions.
Want to find out what your customers are really thinking? Check out Eyes4Research’s state of the art emotion AI and facial recognition technology to discover the impact of your brand’s print, videos, and digital ads. Learn more or request a demo by visiting www.eyes4research.com.
January 13, 2021
Eyes4Research
Market research offers decision-makers a glimpse into consumer behaviors and attitudes toward data-driven strategies. From surveys, polls, focus groups, and even interviews, there are many ways to collect insights to better understand target audiences. When considering testing approaches, knowing the difference between implicit and explicit testing remains imperative to align research initiatives with overall business goals. With this, Eyes4Research has created a guide to help researchers compare and contrast testing methods for optimal research results.
Implicit Testing
Implicit testing, otherwise known as System 1, captures the unconscious responses of consumers. This helps researchers unmask the emotional impact and effectiveness of brand campaigns across a variety of mediums. With most of the decision-making process lying within the subconscious, implicit testing helps identify unintentional judgments in a non-research setting for more effective business decisions.
To capture subconscious perceptions, market researchers have begun using AI and facial recognition technology. These research tools closely analyze microexpressions and eye movements when exposed to brand collateral. The use of emotional analytics helps quantify consumer subconscious behaviors, going beyond the verbal responses of participants. This provides insight into the unconscious decision-making process of consumers and its correlation with campaign materials. Other methods include implicit association testing (IAT), associative priming, and semi-implicit testing.
Implicit testing has many advantages. For one, this method allows researchers to analyze cognition/attitudes without the need for time-consuming introspection. Because this method takes less time, more people are willing to participate, helping researchers obtain insights from traditionally hard to reach groups. Also, participants are less susceptible to response biases, such as social desirability bias, as respondents do not have time to adapt their behavior. These tests can also be conducted from smart devices, giving studies exceptional reach and making them a cheap and easy option for researchers. Most of these studies take place in organic environments to capture authentic reactions. This eliminates the influence of external stimuli for more reliable results.
In terms of disadvantages, implicit testing does not replace explicit methods. Though implicit measures provide valuable insight into consumer subconscious behavior, it does not replace explicit methods. Instead, researchers should utilize both implicit and explicit testing to fully understand drivers of consumer behavior. Implicit testing limitations may also result from the time it takes respondents to comprehend materials. When conducting a study, research professionals must find the appropriate amount of time for participants to process material while capturing initial reactions. This may be extremely difficult, as comprehension times differ from person to person.
Explicit Testing
Explicit testing, also referred to as System 2, analyzes conscious behaviors, giving respondents time to think through their responses and develop a deliberate and logic-based answer. Explicit testing may take the form of a survey, focus group, or any research method that is propositional in nature. This method analyzes factors of purposeful decision-making, allowing respondents to understand and further explain their behavior.
Explicit testing provides more in-depth insights on complex research questions than implicit methods allowing participants to identify the origins of their answers and provide details needed for informed decisions. Similar to implicit testing, explicit methods can be conducted in-person or from electronic devices, allowing researchers to stay within their timeframe and budget. Explicit testing allows participants more control and time to analyze their answers, giving researchers reasoning behind an immediate response.
Explicit testing has several limitations researchers must keep in mind while conducting studies. For one, respondents are susceptible to bias when participating in studies that use explicit methods due to concerns over self-presentation resulting in dishonest responses. These tests also limit the participant’s ability to fully express their feelings, as they do not account for subconscious attitudes. Explicit testing may also be more time-consuming than implicit testing. With more detailed questions, participants may have to take more time to process materials and develop an answer.
Overall, implicit and explicit testing should be utilized in tandem to provide a full view of the emotions and logic that drive consumer behavior. By understanding the key drivers of consumer behavior, brands can optimize their practices to be more impactful.
Looking to better understand your target audience? Check out Eyes4Research’s AI and facial recognition technology to determine which print, video, or digital ads resonate with consumers. Visit www.eyes4research.com to learn more.
January 5, 2021
Eyes4Research
From society and culture, news, politics, health, and even comedy, there seems to be a podcast for every interest. Over the past two decades, podcasts have grown in popularity. In 2019, 55% (155 million) people in the US listened to a podcast, according to Infinite Dial 20. With a myriad of listening options and topics to choose from, podcasts continue to transform the landscape of audio content, creating an outlet for entertainment, information, and creativity.
The Emergence of Podcasting
Podcasting, previously referred to as audio blogging, emerged in 2004 with Apple adding a podcast directory to iTunes, allowing listeners to subscribe and download podcasts over the internet to their devices. Podcasting hit the mainstream with the release of the iPhone in 2007 and the development of Apple’s podcast app in 2012. These technological advancements helped bring in new listeners from all over the world, helping broaden the demographic of podcast enthusiasts. Today, podcasts are produced by one of five segments: media companies, podcast production companies, independent podcasters, non-media businesses/nonprofits, and hobbyist creators.
Who is Listening?
As podcasts make their way into the mainstream, audiences continue to grow and diversify. According to MusicOomph, 51% of podcast consumers are male, while 49% are female. 40% of podcast listeners are between the ages of 12-24, with 39% between 25-54, and 17% 55 and older, as reported by Edison Research. This report also shows that the majority of monthly podcast consumers 18 and older hold some form of a degree with 28% having some grad school or advanced degree, 25% having a 4-year degree, 27% completing some college, and 20% having a high school diploma or less. An article published by Convince & Convert shows that the majority of podcast listeners are affluent, with 45% having an average annual income of more than $75,000. Podcast listener diversity has increased over the past ten years, as data collected by Edison Research shows that 59% of podcast listeners are white (down from 73% in 2008), with African Americans making up 12% of listeners, 11% identifying as Hispanic/Latino, 7% Asian, and 11% refusing to answer or identifying as other.
In terms of how people consume podcasts, a study by Andressen Horowitz claims that Apple Podcasts remains the dominant app for listening, despite declining from 80% to 63% over the past few years. Spotify comes in second, making up 9.5% of all podcast streams, while 2.9% of listeners use Overcast. When looking at where people consume podcasts, this study shows that 48% of consumers listen from home, while 26% of listeners tune in from their car, 12% from work, 4% at the gym/while exercising, and 10% from other locations.
Monetization of Podcasts
Despite its rapid growth in popularity, podcasts remain an under-monetized form of media. To make a profit, podcasters may collaborate with advertisers, sell their content as a subscription, release premium content, attend conferences and events, expand into other forms of media, or even use crowdfunding. Though there are several ways to monetize content, ad revenue is much smaller than other forms of media, monetizing $.01 per listener hour compared to $.11 for radio and $.13 for television, as stated by WNIP. Though podcast revenues were down 19% in June 2020 due to the pandemic, experts forecast podcast advertising revenues to grow 14.7% year over year, reaching almost $1 billion by the end of the year, according to AdExchanger.
The Future of Podcasting
As the amount of listeners continues to grow, the future of podcasting looks promising. Statista forecasts that the number of podcast listeners will increase around 20 million each year, surpassing 160 million in 2023. Podcast episode creation has also increased, with a 27% increase in podcast episodes from 2018-2019, according to TechCrunch. With an increase in ad spending, growth in international popularity, and an influx of celebrity shows, the popularity of podcasting shows no signs of slowing down.
January 4, 2021
Eyes4Research
From online, in-person, to hybrid discussions, conducting a focus group is a great way to collect qualitative data directly from consumers. This research method helps businesses better understand their target audience’s behaviors and opinions for data-driven decision making and the creation of effective strategies. Here are four tips on how to get the most out of your focus group for better results.
Be Intentional with your Questions
The effectiveness of a focus group relies on the questions asked to participants. When developing a discussion guide, researchers must focus on the overall objectives of the study. This ensures each question aligns with the specific research goals of the organization. All discussion questions should be open-ended, using unbiased language that all participants can understand. Each question should be relevant to the research topic and organized to encourage a natural flow of discussion. Though the appropriate number of questions relies on the discussion topic, researchers should aim to keep focus groups under two hours to avoid burnout or boredom by participants.
Identify and Adapt to Participant Behaviors
Running a successful focus group requires vigilance, patience, and adaptability. Though participants are chosen using research objectives, moderators cannot predict the personalities of panelists. Without proper moderation, focus group dynamics may become hostile, compromising overall data quality. When conducting a focus group, facilitators must make participants comfortable and set expectations for communication. To make panelists comfortable, moderators should introduce participants to each other using name tags, icebreaker questions, and even games. Creating a casual rapport promotes discussion while providing the facilitator insight into the best approach for effective moderation.
Focus group leaders must ensure that all panelists can share their opinions. To balance participation, moderators must directly engage quieter respondents while politely acknowledging and limiting overtalking and interruptions by dominant panelists. Moderators must carefully analyze the verbal and nonverbal behaviors of participants early on in the group. This may help focus group leaders adapt their approach to accommodate each member accordingly and make each panelist comfortable. Researchers can do this by paying attention to where members sit in the group, one’s initiative to answer questions, and overall body language.
Recognize your Bias
As a moderator, personal bias can have a considerable effect on the quality of insights. With this, it remains imperative that researchers remain neutral throughout the focus group discussion to prevent swaying opinions and maintain a comfortable space to share insights. Before conducting a focus group, moderators should identify any possible cognitive shortcuts, ensuring questions are written using unbiased language. Interview questions should be open-ended and phrased in a way that avoids leading respondents towards a specific answer. When communicating with participants, facilitators must remain impartial, avoiding verbal and nonverbal cues of agreement/disagreement with opinions. The data collected throughout the focus group should be analyzed with possible biases and study objectives taken into consideration.
When in Doubt, Ask for Clarification
Qualitative research remains an exceptional tool for collecting in-depth, personal insights directly from the source. Despite this, moderators run the risk of collecting insufficient data resulting from a lack of specificity by participants. To prevent moderators from misinterpreting responses, one should always ask for clarification when faced with uncertainty. Moderators can use several tactics to ensure their interpretation of a panelist’s response is accurate. One of the most common tactics consists of the researcher paraphrasing the panelist’s response to them, allowing the source to confirm or correct the interpretation. Another way to probe information is to summarize key points at the end of the focus group, allowing participants to add to their responses. Finally, moderators can simply ask what the participant means regarding specific parts of their response. Whatever the preferred method, requesting clarification from panelists may simplify data interpretation by eliminating vagueness that may result in low-quality data and poor decision-making.
Looking for panelists for your next focus group? Eyes4Research provides clients with B2B, B2C, and specialty/niche panels built and managed by our in-house team of research professionals for high-quality insights. Learn more or request a demo at www.eyes4research.com.