Will shoppers pay for your next food idea?
Describe your product and the new claims you are considering. Palato tests them on up to a thousand synthetic consumers shaped on your market, with research-grade analytical methods drawn from the latest international consumer-science literature, and tells you which novelties matter, what they are worth, and who cares most — before you spend on fieldwork.
The methodology powering Palato is published in peer-reviewed scientific literature.
Food Quality and PreferenceElsevier · 2026 · ScienceDirectThree steps, about an hour
Describe the product
Name, pack, today's price, and the novelties you want to test — organic, origin, a claim, a new pack. Pick from the library or write your own.
Choose your shoppers
Set the profile of your market: gender, age, income, region, habits. Palato draws hundreds of synthetic consumers that match it.
Read the report
Each consumer makes 8 choices and bids in an auction. You get importance, willingness to pay, the most interested segments and a what-if simulator.
A report you can take to the next meeting
Four views of the same consumers, with confidence intervals, and all the data behind them.
What drives the choice
Share of the decision driven by each feature, price included.
What each novelty is worth
Willingness to pay from the choices and price premium from the auction, side by side.
Who cares most
Premium by consumer segment, so you know whom to target first.
What if…
Move the price and the features of your new product and watch its share against today's product and no purchase.
Illustrative numbers for apricot jam. Show another product →
Start free. Pay only for the test you order.
Every test starts with a free trial on 5 consumers, to check your setup and preview the report. Prices are per test, VAT excluded.
- 5 consumers
- Setup check and report preview
- A preview of your report
- Values, segments and downloads
- 200 synthetic consumers
- Up to 3 new features
- What drives the choice and value in € (choices + auction)
- Indicative segments
- Market simulator and scenarios
- Editable Word report with high-quality charts, raw data (CSV)
- 500 synthetic consumers
- Up to 5 new features
- What drives the choice and value in € (choices + auction)
- Segment analysis, map and full table
- Market simulator and scenarios
- Editable Word report with high-quality charts, raw data (CSV)
- 1,000 synthetic consumers
- Up to 5 new features
- What drives the choice and value in € (choices + auction)
- Segment analysis, map and full table
- Market simulator and scenarios
- Editable Word report with high-quality charts, raw data (CSV)
- Priority support: a reply within 24 hours
We don't ask what people like. We watch what they choose.
Ask shoppers whether they want organic, local or low-sugar and almost everyone says yes. At the shelf, the same people compare products side by side, look at the price and give something up. Palato recreates those moments, so the answers reflect the trade-offs people really make, not their good intentions.
At the shelf: choosing between options
Each consumer stands in front of the shelf eight times. Every time there are two versions of your product — different prices, different claims — and the option of buying neither, exactly as in a store. Nobody is asked to rate “organic” on a scale: they simply pick.
Because price and features change together from one shelf to the next, the pattern of choices reveals how much each feature weighs in the decision and how many euros it is worth.
Discrete choice experiment with an efficient (D-optimal) design, analysed with random-utility models — the approach recognised with the 2000 Nobel Prize in economics (D. McFadden).
At the till: putting a price on it
Then each consumer states the most they would pay for today's product and for each new version. A selling price is drawn at random: if their offer is at least that price, they buy — and pay the drawn price, not their offer.
With these rules, exaggerating or understating can only backfire, so the best move is to state one's true value. The difference between the offers for the new and the current product is the premium the new feature can earn.
Experimental auction with the Becker–DeGroot–Marschak mechanism, a standard tool of experimental economics for measuring willingness to pay.
Two lenses on the same person
The same consumer goes through both situations. When choosing at the shelf and naming a price at the till point to the same value, you can rely on it more. When they disagree, it is a signal worth investigating before you invest.
From thousands of answers to a decision
Each synthetic consumer has a profile drawn from your market — age, income, region, habits — so the answers can be split by group. You see which features matter most, what each is worth, which segments care the most, and how your share would move if you changed price or features.
Every test is reproducible: model, settings and every question are saved, and the raw data come as CSV files, ready for any statistics software.
The science behind Palato
Palato is built on the methodology validated by our team in Food Quality and Preference (2026). Read our paper on how synthetic samples replicate human choice experiments:
Do large language models shop like people? Comparing GPT and human food choices using discrete choice experiments ↗Palato brings that research experience into a practical tool for exploring product ideas and preparing consumer tests.
What Palato is — and is not
The consumers are synthetic: a large language model answers as a person with a given profile would. Our own study and other recent research on these “silicon samples” finds that they often reproduce the direction and ranking of real preferences, but with less variety than real people, more stereotyped segments and some sensitivity to wording.
Use it to screen ideas and to design the real test: which claims to keep, which price range to try, which segments to recruit. Do not use it as a market forecast or as evidence for a claim on pack.
From consumer science to your shelf
Palato is a research prototype that grows out of university work in marketing and consumer science. Every full test is reviewed before it runs; a free pretest on 5 consumers is open to every registered user.