Features Conjoint & MaxDiff Survey Items Pricing & Kano Product Evaluation Analytics Market Simulation Survey Analysis How It Works Transparency Pricing Log in Start Free

Market Simulation

Six questions for your market.
Answered from every respondent’s own preferences.

After estimation, every respondent has their own part-worth utilities. The simulator places them in front of a market you define, your product next to its competitors, and turns them into shares, prices and the value of a change. Each tool comes with a short recording from the app.

  1. Choice Share

    Which product wins, and by how much?

    The share each product would take in the market you set up, with or without the option to buy nothing.

    You getShare per product with a 95% interval

  2. Willingness to Pay

    What is a feature worth in money?

    The price premium a change can carry before the product falls back below the share it had.

    You getSimulation WTP, Robust WTP and Pseudo-WTP

  3. Reservation Price

    Up to which price does each person still buy?

    The price at which your product stops beating the best alternative, person by person.

    You getMedian and mean price limit, demand curve

  4. Price Elasticity

    How strongly does share react to price?

    Share across a price range you choose, and the point where demand turns elastic.

    You getOwn-price and cross-price elasticity

  5. Sensitivity

    Which level of one attribute pays off?

    One attribute, tested level by level, for up to five products at once.

    You getShare for every level, per product

  6. Persona

    What would a person with this profile choose?

    Describe a customer by age, attitudes or habits and read their preferences and choice.

    You getImportance, utilities and choice per profile

One market, six tools

Set up the market once

All six tools work from the same run and the same products panel. Change a competitor once, and every tool calculates with the new market.

The modelWhich run

Choose the estimation run the simulator works from. Hierarchical Bayes gives every respondent their own utilities; an AI-Driven Profile Analysis run adds the profile model that Persona needs.

The productsUp to 20

Describe your product and its competitors with the levels of the study and a price. Attributes estimated as linear, such as a linear price, take any value in between. Save the set, or import a fixed task from the design.

The pricePrice setup and currency

Tell Choicery which attribute is the price and which numbers its levels stand for, and pick the currency the results are shown in.

The peopleFilters

Narrow the sample by up to five survey variables, for example heavy users in one age group. The tools show that segment next to the result for everyone.

The guideOne click from each tool

Every tool links its own methodology guide: the calculation, its assumptions and how to report the numbers.

The hand-overExports and PDF

Each tool exports its result as Excel, CSV or PNG. One PDF report collects the open tool and every other tool with a current result.

Tool 1 of 6 · Choice Share

Which product wins, and by how much?

With the default rule, every respondent gives a choice probability to every product in your market, computed from their own utilities. Choicery averages these probabilities over the sample. The result is the share each product would take, and the share of people who would buy nothing at all.

In the recording: four pizzas compete, one respondent’s probabilities become the average over 300 people, and the intervals show how certain the lead is.

You set

  • The products in the market and their prices
  • The choice rule: Share of Preference as the default, Randomized First Choice to correct for near-identical products, or First Choice, where each person takes only the product they like best
  • Whether the option to buy nothing takes part

You get

  • Share per product and for buying nothing, with the leader marked
  • 95% intervals on request, from 200 resamples of the respondents
  • Individual probabilities for every respondent as a download
  • CSV and PNG of the chart, and the filtered segment next to everyone as Excel

Why the interval matters. In the recording, the leading pizza has 32.8% and the runner-up 30.6%. The two intervals overlap, so the lead is small.

Tool 2 of 6 · Willingness to Pay

What is a feature worth in money?

Change one attribute of your product, for example faster delivery, and its share rises. Choicery then searches for the higher price at which the share is back where it started, with all competitors still in the market. The price difference is what the change is worth in that market.

In the recording: 20-minute instead of 35-minute delivery is worth $1.77 in the configured market; Robust WTP then repeats the calculation in 300 random markets.

You set

  • The focal product and the attribute that changes
  • The level after the change
  • The competitors it has to hold its own against

Three ways to read it

  • Simulation WTP: the break-even premium in the market you configured, plus the median and mean across respondents
  • Robust WTP: the same calculation in hundreds of randomly drawn markets, reported as a median with a bootstrap interval, so the answer does not hinge on the competitors you happened to pick
  • Pseudo-WTP: the classic ratio of utility difference to price coefficient, quick and without competitors, for one attribute change or a whole product against another

Which one to report. Robust WTP is the recommended figure, because a single market can flatter or understate a feature. Simulation WTP and Robust WTP work from a price estimated as a linear attribute.

Tool 3 of 6 · Reservation Price

Up to which price does each person still buy?

For every respondent, Choicery calculates the highest price at which your product still beats the best alternative that person has: a competitor or buying nothing. The price limits of all respondents give the typical limit and a demand curve.

In the recording: the pizza costs $11.00 today, the median price limit is $10.72, and at $11 just under half of the people would still choose it.

You set

  • Your product, as the first product in the panel
  • The competitors it is measured against, as the other products

You get

  • Median and mean reservation price, the mean with a 95% interval
  • A demand curve: the share of people who would still buy at each price
  • Every respondent’s price limit in the Excel export, next to the summary and the curve
  • The filtered segment next to the result for everyone

People who never buy stay in. Respondents who prefer an alternative even at a price of zero are counted at $0 and kept in the sample, so the median does not look better than the market is.

Tool 4 of 6 · Price Elasticity

How strongly does share react to price?

Choicery simulates your product at a series of prices between a minimum and a maximum you set, while the competitors keep theirs. From neighbouring prices it computes the elasticity, so you see where a price increase starts to cost proportionally more share than it adds in price.

In the recording: two simulations give one arc elasticity; ten prices give the whole curve, which turns elastic just above today’s price of $11.

You set

  • The focal product
  • The lowest and the highest price to test
  • The choice rule and the option to buy nothing

You get

  • Arc elasticity across the whole range, and an average elasticity from a log-log fit with its R²
  • The demand curve with the elasticity at each price, as CSV and PNG
  • Cross-price elasticity: how your share moves when a competitor changes its price
  • Elasticity for the filtered segment, and an Excel export

How to read −1.2. One percent more on the price costs about 1.2 percent of the share. Beyond 1 in absolute terms, demand is elastic. If price was estimated in levels, Choicery interpolates between them.

Tool 5 of 6 · Sensitivity

Which level of one attribute pays off?

Pick one attribute, for example delivery time. Choicery runs one simulation for every level of it, changing only that one product while everything else stays as you set it. You see at a glance which level wins share, which one costs it, and where two products swap places.

In the recording: 20-minute delivery adds 5.65 points for the pizza; 65 minutes would cost almost 15.

You set

  • The attribute to vary
  • Up to five products to plot, while every product in the panel keeps competing
  • The choice rule and the option to buy nothing

You get

  • Share for every level, product by product, with the change against today
  • Crossover points where the ranking of two products flips
  • For price, two views: each product moves its price alone, or all of them move together
  • The filtered segment next to everyone, CSV and PNG

Each curve is its own market. Curves at the same level do not add up to 100%, because each one comes from a separate simulation in which only that product changes.

Tool 6 of 6 · Persona

What would a person with this profile choose?

An AI-Driven Profile Analysis run learns how preferences depend on what you asked about the respondents, such as age, attitudes or habits. Persona uses that model the other way round: describe a profile, and Choicery predicts the preferences and the choice of such a person, even if nobody with exactly that profile took the survey.

In the recording: a student picks the focal pizza with 50.2% and cares most about price; a vegetarian family picks a different one.

You set

  • The profile, from the variables your Profile Analysis run was trained on
  • Up to five profiles to compare side by side

You get

  • Attribute importance and part-worth utilities for the profile
  • Product preferences: how such a person would split their choice across your market
  • Persona comparison: profiles in one table, with different winners where their tastes differ
  • CSV and PNG downloads, and a place in the PDF report

Built on a Profile Analysis run. Persona needs the profile model that the AI-Driven Profile Analysis or Neural HB estimates. The sliders stay within the range observed in your sample.

Put your own market to the test

Run a study, estimate it, and every tool on this page works with your data. The free trial includes the market simulator.

Start Free See the estimation models

Explainer