Why Conjoint Models Can Misrepresent Real Market Share

Executive Summary

Many conjoint studies assume that all brands compete equally, features are fully understood, and price operates independently of quality perception.

In real markets, buyers filter first — by awareness, eligibility, distribution, relationships, and risk.

When those filters aren’t built into the design, projected share can be overstated and smaller brands can appear more viable than they truly are.

This article outlines the three most common design problems that distort conjoint results — and how to fix them:

1. The Equal Awareness Problem
2. The Feature Inflation Problem
3. The Price–Quality Problem

Conjoint is powerful — but only when it reflects how buyers actually move from awareness to shortlist to choice.

---

Many conjoint studies look accurate — but miss how real buying decisions unfold.

When a model assigns meaningful share to brands that would never make a real-world shortlist, confidence erodes quickly.

This isn’t a math problem. It’s a design problem.

In actual markets, buyer decisions are shaped early—by awareness, eligibility, distribution, relationships, and risk—long before features are compared.

That assumption alone can materially distort projected share.

---

Why Conjoint Models Often Misrepresent Real Market Outcomes

Conjoint models estimate preference—but real-world outcomes are shaped by more than preference.

In actual markets, outcomes are determined by:

• Awareness: whether buyers know the option exists
• Consideration: whether the option is seen as viable
• Credibility: whether the option is trusted
• Risk: whether the option feels safe to choose
• Constraints: availability, relationships, and switching barriers

Because most conjoint models assume equal awareness, full consideration, and independent evaluation, they can systematically misrepresent real-world outcomes—even when the model appears statistically sound.

This is not a statistical issue—it is a structural one.

---

1. The Equal Awareness Problem

If a brand appears in a choice task, respondents treat it as a legitimate option.

But in the real world:

• Low-awareness brands are ignored
• Poorly distributed brands never make procurement lists
• Incumbents benefit from switching inertia
• Vendor approval policies limit eligibility

If consideration is not measured and built into the forecast, minor brands can receive artificially high simulated share simply because respondents were allowed to choose them.

Projected share should reflect two things:

• The likelihood of being considered
• The likelihood of being chosen

Not just the second.

How to fix it:
Measure awareness before trade-off tasks. Capture brands considered in the last purchase. Ask about likelihood of consideration in future purchases. Share forecasts should reflect both entry into the competitive set and performance within it.

---

2. The Feature Inflation Problem

A second distortion occurs when features are explained more clearly in the survey than they are in the market.

Inside many studies:

• Benefits are described in detail
• Technical features are simplified
• Performance advantages are clearly spelled out

In reality:

• Sales messaging varies
• Buyers misunderstand features
• Some benefits go unnoticed

When this happens, we start educating respondents rather than capturing how they normally decide.

The survey presents the product in its best light, making certain features appear more powerful than they are in actual buying situations.

How to fix it:
Use market language. Mirror how competitors describe features. Avoid strengthening the value proposition beyond what buyers typically encounter. The goal is to measure preference — not improve it during the study.

---

3. The Price–Quality Problem

Conjoint assumes attributes operate independently. In theory, lower price increases attractiveness.

In practice, price signals quality.

An unrealistically low price can trigger rejection — not because buyers dislike savings, but because it feels suspicious. If price ranges extend beyond real market boundaries, elasticity estimates can become unstable.

How to fix it:
Keep price ranges realistic. Avoid extreme low-price scenarios unless they truly exist in the market. Test whether unusually low prices are driving rejection for the wrong reasons.

Prices shown in the study should match what buyers would actually encounter.

---

The Bigger Issue

Most conjoint studies model only the final step: comparing features.

Real purchase decisions typically involve:

1. Awareness
2. Consideration filtering
3. Risk assessment
4. Budget constraints
5. Feature trade-offs

If the earlier stages are ignored, the final stage becomes inflated.

A model can be technically correct and still not reflect real-world behavior.

Before relying on projected share forecasts, leaders should ask:

• Was brand consideration measured and incorporated?
• Were awareness differences built into the forecast?
• Were features described as they are in the real market?
• Were the prices shown realistic?

If not, opportunity may be overstated — and capital misallocated.

---

Final Thought

Conjoint is a powerful strategic tool.

But clean numbers don’t guarantee real-world accuracy.

Only when the design mirrors how buyers move from awareness to shortlist to choice does conjoint produce forecasts leaders can trust.