Capturing nuance: The end of the Likert scale

Developed by social psychologist Rensis Likert in 1932, the Likert scale was designed to capture subtle nuances and simplify data collection and allowed for easier statistical analysis of attitudes.

With a spectrum of answers to choose from, it’s still being used today in market research surveys to collect nuanced data— it’s an uncomplicated, reliable way to get the insights brands need to make major go-to-market decisions in an increasingly fast-paced world... right?  

Imagine being at a dinner party, surrounded by friends old and new, conversation flowing naturally between all in attendance. People are sharing their thoughts on current events and pop culture, sometimes agreeing to disagree, but the vibe of the night nurtures the sharing of unfiltered takes and nuanced opinions. Imagine that while at this dinner party, you enter a conversation with someone about a viral ad campaign that’s been the subject of much debate on social media. You ask your friend what they think, and they answer with “I was moderately amused” and stop there. You might try for more nuance and ask them what they thought of the casting – they answer simply, “Not satisfied at all”. You tell them you work for the casting company, and you’ll get them a refill if they tell you what they really thought of it. They respond, “Excellent,” and hold out their cup. You’d probably check their pulse before taking your cheese and crackers into the next room in search of something more stimulating.   

The answers that researchers get from the Likert scale fail to capture the complexity and nuance that exists today. People don't need to be convinced with freebies or discount codes to share their thoughts on a particular subject in the real world, where brands actually exist. There’s nothing less natural to today’s consumer who’s used to making their thoughts known on social media feeds than putting oneself in a box. Eliciting real answers requires asking real questions in the real world –– the one we’re in now. 

While the attempt was to capture beyond yes or no answers, the Likert scale ultimately still quantifies what isn’t precisely quantifiable, meaning results from Likert scale tests reflect more missed opportunities than valuable insights. If you’re going to get answers that are accurate and capture attitudes in a holistic way –– they’ll contain a multitude of actionable insights instead of compressing them into pre-determined answers by imperfect survey takers. 

The Likert scale survey versus in the wild testing

Capturing real nuance requires recording actual behavior and uninhibited opinions, not asking people to point out their own unique quirks, preferences, or variations in opinion on a limiting scale.   

Likert scale answers are often unreliable because of various biases:  

• Acquiescence bias: people tend to agree with statements, whether they actually do or not.
• Central tendency bias: Some cultures or individuals avoid extremes, clustering answers around the middle. 
• Extreme response bias: The opposite — some people love picking extremes.  
• Cross-cultural comparisons: A “4” on a 7-point scale may mean something different in Japan vs. the U.S.  

In the wild testing yields results that are less impacted by these kinds of biases because test subjects don’t know they’re part of a test. 

3 little words can determine the success of your campaign: We’re talking about the “say-do” gap, a well-documented phenomenon in the world of market research. We just aren’t as good at accurately reporting on our own behavior as we think we are. Despite our best efforts to be honest, we tend to have blind spots.  

For example, most people – 68% of them – claim to prefer their Pop-Tarts toasted, but we did the testing in the wild by getting people talking in-feed about their snacking goals versus reality to see if people are actually taking the time to toast before tasting.  

• 2.32% said they preferred a toasted Pop-Tart 
• 2.91% couldn’t be bothered – they enjoyed right out of the wrapper
• The “say-do” gap: 0.59%
  

Why nuance matters in modern insights

Consumer decisions today are shaped by layers of context — emotion, culture, identity, environment — that can’t be flattened into a five-point response. 

• A “somewhat agree” tells you little about why someone feels the way they do.
• Richer methods — observational data, open-ended responses, and in the wild testing — surface the key drivers behind consumer buying behavior. 
• Nuance isn't just for show; it directly fuels stronger creative, sharper targeting, and faster pivots.

In a recent case study with Allagash Brewing, we tested a variety of taglines speaking to various aspects of their brand identity to see what resonated most and why. On a limited 5-point scale, Likert responses could say they were equally effective, but in the wild testing showed that one tagline pulled way ahead, driving 40% more engagement than the others, because it resonated with the core brand value of community and being a Maine-born brand –– this simple realization led to an entire brand strategy that was more aligned with what their consumers actually liked about the brand.

The future beyond Likert 

Research methods are evolving toward mixed approaches that embrace complexity.

With new technologies ramping up in virtually every other sphere every day, rapid exchanging of information, opinions being formed by algorithms, around the world and in seconds – a 5-point scale developed nearly a century ago just isn’t on the level.  

For example, unbranded and targeted in the wild testing on social feeds lets you go undercover without risking competitors beating you to market or for negative responses to compromise your brand identity, captures answers to questions you may not even have thought of, and fosters the authentic connections that winning brands will make in a landscape where every day is innovated upon. 

The goal of market researchers is to gain insights with the tools of tomorrow, today.

That is: 

• Using AI tools to analyze unstructured, qualitative data at scale. 
• Leveraging targeted, large-scale testing reveals cultural nuance. 
• More dimensionality, not more data points.

What’s more — When traditional survey data, social listening, and in the wild testing are combined, the triangulation reduces reliance on reductive methods of testing, like Likert. 

While a tool like the Likert scale has significantly contributed to the development of capturing social nuance, with the innovations of social media-forward world we’ve built since, there’s richer, more relevant insights available. Testing that occurs in the new real world, on the social feeds where consumers live, reveals what people actually think, feel, and do — in the wild is where you go to develop confident campaigns that you already know will resonate.