Consumer research for food

What is sensory analysis?

Updated September 11, 2026 · 6 minute read

Short answer

Sensory analysis measures how people perceive a food product through the senses — sight, smell, taste, touch and sound. It turns the subjective experience of taste into measurable, usable data, so food companies can make informed decisions about recipe, packaging and positioning before launch. It takes place in a sensory laboratory or, increasingly, in consumers' own homes with video recording and AI analysis.

In this article

What exactly does sensory analysis cover?

Sensory analysis (also called sensory evaluation) is the scientific discipline that measures how people experience a product through the five senses. For food that means taste, smell, texture (mouthfeel), appearance and sound — think of the crunch of a cracker or the snap of a bar. The goal is to turn that experience into reliable data a product team can act on.

The field combines food science, psychology and statistics. Where a chemical analysis tells you what is in a product, sensory analysis tells you how people actually experience it — and whether they would buy it. Those two things do not always line up, and that difference often decides whether a launch succeeds. Peer-reviewed estimates put new food product failure at 50–75% within two years, which is why teams increasingly test perception before committing to production.

What types of sensory analysis exist?

Sensory analysis broadly falls into three categories, each answering its own question:

1. Discrimination testing (are the products different?)

This establishes whether a perceivable difference exists between two or more products — for example after a recipe change or an ingredient swap. The most common methods are the triangle test and the duo-trio test. Useful for reformulation, cost reduction or a supplier change.

2. Descriptive analysis (how do they differ?)

A trained panel describes and rates a product's exact properties on a fixed profile: how sweet, how bitter, how creamy, how crunchy. The result is a detailed “sensory profile” for objectively comparing products against competitors or against a target profile.

3. Affective testing / consumer research (do people like it?)

Here preference and acceptance among the real target audience are what count: do consumers like it, which product do they prefer, and would they buy it? This type sits closest to consumer research for food products and is the most predictive of market success.

Sensory analysis in the lab vs. at home

The classic approach uses a sensory booth: a neutral, controlled space where participants taste isolated samples. It offers control, but it lacks context. An increasingly common alternative is in-home testing: consumers test the product at home, in the situation where they would normally consume it.

AspectLab / boothIn-home
EnvironmentNeutral, controlledReal kitchen, real context
BehaviorIsolated sample tastingPreparing, cooking, eating as at home
Predictive valueHigh for pure taste differencesHigher for purchase intent and usage
TimelineTypically 4–8 weeksOften within 1–2 weeks
CostsHigher (facility, logistics)Lower (consumer's own device)
Important Lab and home do not exclude each other. The lab is strong at isolating a pure taste difference; in-home testing is stronger at predicting whether people will actually buy a product and keep using it. Many teams combine both.

How does a sensory analysis work?

However much the methods vary, almost every sensory analysis follows the same four phases:

  1. Research question and design. Establish what you want to know (difference, profile or preference), which product you are testing and who the target audience is.
  2. Participant recruitment. Select a trained panel (descriptive) or a representative group of consumers (affective) matching your market.
  3. Running the test. Participants taste and rate the product under standardized conditions, in a lab or at home.
  4. Analysis and reporting. The data is analyzed statistically and translated into concrete recommendations for the product team.

For a more detailed practical walkthrough, see how to test a new food product.

How much does sensory analysis cost, and how long does it take?

There is no fixed price: costs depend on the method, the number of products, the panel type and the number of participants. Descriptive analysis with a trained panel is labor-intensive; large-scale consumer research needs more participants. Laboratory analysis is usually the most expensive because of the facility, scheduling and logistics.

On timing, plan for 4 to 8 weeks for traditional laboratory sensory analysis, including recruitment and analysis. Approaches that recruit from an existing consumer network and automate the analysis with AI reduce that to about one week.

Sensory analysis software: what it does, and how AI changes it

Sensory analysis software traditionally covers study design, data collection (scorecards and questionnaires), statistics and reporting. It makes a sensory team faster, but it still measures what participants say — ratings and answers.

A newer generation of tools is video-first. Eatpol, a spin-off of Wageningen University & Research, combines in-home testing with AI video analysis: consumers test a product at home with their own smartphone, and the app records the full journey — from unboxing through preparation to the first bite.

Proprietary computer vision then analyzes bites, chewing, facial expressions, eating speed and spontaneous comments. The subjective experience of taste becomes measurable behavioral signals that correlate with purchase intent — within one week on average, without a sensory booth. Because it measures what people do rather than what they predict about themselves, it also addresses the say-do gap that undermines survey-based methods.

Want to know which method fits your product? Explore our consumer research platform (Vox, Domus and Studio), or use Eatpol Nova to find what you could launch next.

Frequently asked questions about sensory analysis

What is the difference between sensory analysis and consumer research?

Sensory analysis measures the perceivable properties of a product — taste, smell, texture, appearance. Consumer research measures how a target audience responds to those properties: preference, purchase intent and emotion. In practice they are combined: first establish what objectively characterizes a product, then whether consumers like it and would buy it.

How many participants does a sensory analysis need?

It depends on the type. Descriptive analysis with a trained panel often needs only 8 to 12 panelists. Acceptance and preference testing with consumers typically needs 60 to 100+ participants from the target audience to demonstrate reliable differences.

How much does sensory analysis cost?

Costs vary widely with the method, the number of products, the panel type and the number of participants. Classic laboratory analysis is usually the most expensive because of the lab facility and logistics. AI-guided in-home testing lowers the fixed costs because consumers test with their own smartphone in their own kitchen.

How long does a sensory analysis take?

Traditional laboratory sensory analysis often takes 4 to 8 weeks including recruitment, scheduling and analysis. With Eatpol's approach — recruiting from an existing consumer network and analyzing video with AI — results are available within one week on average.

Can sensory analysis be done at home instead of in a lab?

Yes. In-home sensory testing has consumers test the product in their real environment, where food decisions are actually made. That produces more realistic behavior than a sensory booth. Eatpol records everything on video and analyzes bites, chewing, emotions and spontaneous comments with computer vision.

What is sensory analysis software?

Sensory analysis software collects and analyzes sensory test data: study design, participant responses, statistics and reporting. A newer generation adds AI video analysis — instead of only collecting scores, it records consumers testing at home and extracts behavioral signals such as bites, chewing, facial expressions and spontaneous comments. Eatpol's platform is an example of this video-first approach.

Ready to see your product through consumers' eyes?

Eatpol runs AI-powered in-home consumer research for food brands — real bites, emotions and purchase intent, within one week on average.

Book a 30-minute call → Explore the platform