Consumer Insight Triangulation
Combine stated feedback, observed behavior, and performance data
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 98%
Consumer Insight Triangulation combines multiple imperfect signals to decide what to develop next. Begin with the consumer need or pain point that the brand intends to address. Gather qualitative feedback through interviews, focus groups, or prototype exposure, but also observe behavior that participants may not articulate: laughter, hesitation, attention, affection, or withdrawal. Add quantitative evidence such as ratings, dial tests, pilot results, or digital A/B tests. Then account for distortions, including dominant focus-group participants and the limits of small samples. The final decision comes from themes that recur across methods, interpreted through creative and commercial judgment. Research informs the work rather than mechanically deciding it, preserving the productive balance Levy describes as art and science.
Origin
Extracted from Masters in Business, where Sarah Kirshbaum Levy described Viacom and Nickelodeon's layered research across focus groups, behavioral observation, ratings, pilots, dial tests, and digital experiments.
Core principles
- 01Start with a real consumer need or pain point
- 02Treat every research result as an input
- 03Combine qualitative and quantitative evidence
- 04Observe behavior as well as words
- 05Preserve creative judgment alongside data
How to run it
- 1
Frame the consumer question
Define the need, pain point, reaction, or decision the research must illuminate. Tie the question to a specific product, message, character, or experience.
Pro tip Ask what evidence would change the team's next action.
Watch out Broad questions such as “Do consumers like us?” produce little actionable information.
- 2
Collect qualitative signals
Use interviews, focus groups, and prototype exposure to hear consumers' language and identify recurring themes.
Pro tip Run multiple groups so one unusual room does not become the whole answer.
Watch out A loud participant can alter what everyone else says.
- 3
Observe authentic behavior
Watch what people do while interacting with the material, including attention, laughter, avoidance, or physical engagement.
Pro tip Behavior can be especially revealing when participants lack the vocabulary to explain their response.
Watch out Do not infer a universal conclusion from one dramatic reaction.
- 4
Add quantitative evidence
Measure performance through ratings, pilot tests, surveys, dial testing, or controlled digital experiments.
Pro tip Match each metric to the consumer question rather than collecting data indiscriminately.
Watch out Post-launch metrics explain what happened more readily than why it happened.
- 5
Triangulate the evidence
Look for conclusions supported by several methods and identify where signals conflict. Select the next iteration or test rather than treating the findings as immutable truth.
Pro tip Use disagreement between methods to design a more discriminating follow-up test.
Watch out Do not let research override strong creative instincts without understanding the cause of the conflict.
In the wild
A studio shows several character designs to children in separate groups. Researchers record comments but also observe which designs make children smile, lean forward, or reach for the toy. A pilot scene and dial test then reveal whether attention persists during the story. The team advances the character supported by both spontaneous behavior and measured engagement.
→ The studio chooses a character with evidence of immediate appeal and sustained attention.
A financial app interviews prospective customers about account-opening anxiety, watches users complete two prototypes, and A/B tests the strongest designs. Interviews explain the hesitation, observation reveals where users stall, and conversion data shows which revision works at scale.
→ The team ships a simpler flow backed by converging qualitative and quantitative evidence.
Common mistakes
Obeying one focus group
A small group can be distorted by social dynamics, sampling, or one forceful participant and should be treated as one input.
Listening only to words
Consumers may be unable or unwilling to articulate their authentic response, while their behavior provides a clearer signal.
Replacing judgment with metrics
Data cannot fully substitute for the instincts of experienced creators or explain every exceptional idea.
Is it for you?
Best for
It is best for teams developing brands, products, characters, content, or messages for a defined audience.
Not ideal for
It is not ideal when leaders intend to use research as a substitute for judgment or to justify a predetermined answer.
From the transcript
“So, I mean there's so many levels, you know, both quantitative and qualitative.”
“We were basically bringing things to kids and watching them react.”
“But again, all of this was in service of asking the right questions of your consumer and getting to them in kind of different ways…”
From the episode
Sarah Kirshbaum Levy on the Value of Marketing