How do you do market research for a food or beverage product?
TL;DR
- Research the idea, the substitution, the price and the name before the sensory test, then taste only the finalists.
- New products fight habits. Buyers tend to overvalue what they already buy, so a new product must be clearly better, not just a bit better.
- Read results by heavy and light category buyers. They want different things.
- Neon Blue runs concept, price and name reads by shopper segment, labeled as simulated. Run your first study free
Neon Blue builds simulations of human behavior. Food and beverage brands research a new product in a set order: is the flavor or concept wanted, what does it replace in the basket, what should it cost at the shelf, and what should it be called. Taste itself still needs real mouths, so sensory work happens with real people. Everything before and around the taste test, from which flavors to make to which price to print, can be read on a simulated audience of category shoppers first, so the expensive sensory round tests fewer, better options.
Why do so many new food and beverage products struggle?
In 2006, John Gourville of Harvard Business School described what he called the 9x effect in the Harvard Business Review. Consumers tend to overvalue the products they already use by about a factor of three, and companies tend to overvalue their own new product by about the same. The result is a mismatch of roughly nine to one between what the company thinks the new product is worth and what buyers do.
Grocery is full of habit. People reach for the same yogurt, the same soda, the same chips every week. Research for a new food or beverage product has to measure not just "do you like it?" but "would you switch from what you buy now?" That second question is where most of the 9x gap shows up.
What should food and beverage research cover?
| Stage | Decision | Study | What to read |
|---|---|---|---|
| Idea | Which flavors or concepts to develop | Concept screen | Appeal, uniqueness, fit with the brand |
| Basket | Where volume comes from | Substitution read | What it replaces, new versus cannibalized |
| Price | Pack price and size | Price ladder | Intent by price, pack size choice |
| Name and pack | What to call it | Name test | Appeal, fit with flavor, memorability |
| Taste | Is it good enough | Sensory test with real people | Liking, intent after tasting |
The concept stage uses the questions in what to ask in a concept test. The price stage uses a willingness to pay study.
What questions should a food or beverage concept test ask?
- How often do you buy [category]? [Never to more than weekly]
- Which brands or flavors do you buy most often?
- How appealing is this new flavor? [1 to 5]
- If you bought it, what would you buy less of? [Current products, plus "it would be extra"]
- At $[price] for [pack size], how likely are you to buy it? [1 to 5]
- When would you eat or drink it? [Occasions list]
- What is the one thing that would stop you trying it?
Question 4 tests the 9x effect directly. If most answers are "it would replace another of your products", the launch may shift volume rather than add it. Question 6 points to where it sits on shelf and in ads.
What does a food or beverage study look like on a simulated audience?
Hypothetical: a snack brand has six flavor ideas for a new protein chip and budget to make sensory samples of only two.
- Audience: US adults who bought chips or protein snacks in the past month, split by heavy and light snack buyers, by those who track protein, and by households with and without children.
- Design: monadic concept read for each flavor, substitution question, then a price ladder on the top two at two pack sizes.
- Readout: appeal and intent by flavor and segment, what each would replace, and the price where intent drops.
Example output layout (illustrative, not a real result)
| Flavor | Appeal (top 2) | New volume vs replaces own brand | Intent at single-serve price | Main barrier (coded) |
|---|---|---|---|---|
| Flavor A | __% | __% / __% | __% | Coded theme |
| Flavor B | __% | __% / __% | __% | Coded theme |
| Flavor C | __% | __% / __% | __% | Coded theme |
The two flavors that win on appeal and bring new volume go to sensory testing with real people. A flavor that wins on appeal but mostly replaces the brand's own chips is a weaker launch than the scores suggest.
What do food and beverage teams need to watch for?
- Taste beats talk. A concept that wins on paper can lose after tasting. Treat simulated reads as a filter, not the final answer.
- Claims need care. Health and nutrition claims carry legal rules. Test how buyers read a claim before it goes on pack, and get legal sign-off separately.
- Regional tastes differ. Read results by region if the launch isn't national.
- Keep sources separate. Simulated reads and sensory panels go side by side, never blended.
Restaurant brands have related but different questions, covered in how restaurant brands research guests.
Questions about food and beverage market research
How do you test a new food product before launch?
Screen concepts and flavors first, read what the product would replace and what it should cost, then run sensory tests with real people on the finalists only.
Can you research a beverage without making samples?
You can research the concept, flavor names, pack, price and occasions without samples. Taste itself needs samples and real people.
How many flavors should a concept test include?
As many as you're seriously considering, often five to eight in a screen. Narrow to two or three for sensory work.
How does Neon Blue help food and beverage brands?
Neon Blue builds a simulated audience of your category shoppers and runs concept, substitution, price and name reads by segment, labeled as simulated. Start a free first study.