Keep a conjoint scenario separate from a market forecast

Write down the market inside the simulator

List the exact offers in the scenario, their attribute levels, prices and any none-of-these option. Confirm which customer population the responses represent. A share cannot be interpreted without knowing what it is a share of. If a relevant competitor or route to doing nothing is absent, the simulated choices answer a different question from the actual buying environment.

Sawtooth Software warns that conjoint preference shares generally should not be read as market shares. Its documentation notes that awareness, distribution, stock availability and other real-world influences may be outside the model. This supports using the output as a conditional scenario unless a separate, justified forecasting process supplies the missing context.

Separate a valid change from an unsupported extrapolation

Ask the analyst which attribute combinations and ranges the study can support. A slider may permit an extreme price or a feature combination that respondents never meaningfully evaluated. Technical ability to enter a value is not evidence that the model predicts it reliably. Keep infeasible combinations visibly excluded rather than displaying them as attractive launch options.

Change one decision-relevant element at a time when the aim is to understand its contribution. For broader scenarios, state every changed assumption together. Save the baseline and scenario configuration so another analyst can reproduce the comparison. A screenshot of the final percentage loses the assumptions that produced it and makes later stakeholder retellings difficult to correct.

Challenge the route from preference to business outcome

Map the steps a real customer must pass: learn about the offer, find it available, understand the terms, choose it and complete the purchase. Ask which steps the research actually measures. Do not insert convenient conversion multipliers merely to make the simulator agree with a desired commercial target.

If a forecast is required, give a qualified analyst a separate brief for external evidence, calibration, uncertainty and validation. Keep the original preference model visible beneath any adjustments. The business should be able to tell which information came from respondents, which came from market records and which remains a planning assumption. That separation makes sensitivity analysis useful rather than decorative.

Use the scenario to choose a bounded next step

Compare plausible scenarios and identify decisions that remain robust when assumptions change. The best next action may be to refine a feature package, investigate a competitor or run a limited market test. A narrow, well-supported choice is more useful than an exact-looking sales number built from an incomplete environment.

Place a plain-language interpretation beside every reported percentage: population, alternatives, time of research and important omissions. Record whether the result is preference, a calibrated forecast or observed sales. Revisit the scenario when the competitive set, distribution plan or product specification changes. Do not retain an old model output as a fixed promise simply because it once appeared in an approved deck.

Make the reasoning concrete.

A fictional software company receives a conjoint scenario in which its proposed plan leads three listed alternatives. Sales planning treats that simulated share as next year’s market share, although awareness and sales coverage were not modeled.

The team uses the scenario to compare plan configurations. It labels the result as conditional preference and keeps awareness, distribution and adoption assumptions outside that measured finding.

Take the questions into your next brief.

Use these prompts in your own notes. Nothing is submitted from this page.

1. Alternatives

   Which offers and nonpurchase choices are included?

2. Population

   Whose responses support the estimate?

3. Range

   Which scenarios stay within defensible model use?

4. Omissions

   Which real-world purchase conditions are unmodeled?

5. Decision

   What action is supported without pretending the simulation is observed demand?

Watch for the wrong conclusion.

• Reporting preference share as market share without qualification.
• Exploring impossible product combinations and treating them as recommendations.
• Changing scenario assumptions without preserving the baseline.
• Tuning an output to a target and then calling the target independently validated.

Sources & scope

Workflows, worksheets and hypothetical examples are original editorial analysis. Platform documentation supports only identified technical principles. No measured business result or named human reviewer is claimed.

• Sawtooth Software: Introduction to the Choice Market Simulator →Primary research-software provider methodology guidance · Supports: Preference-share interpretation and real-world influences not automatically captured by conjoint utilities. Scenario review and decision controls are original editorial analysis.