MaxDiff and conjoint analysis solve different research problems. MaxDiff identifies which features, messages, or attributes matter most to respondents by forcing best/worst choices from item sets — producing clear importance rankings. Conjoint analysis goes further: it evaluates full product profiles to measure how respondents trade off attributes against each other, revealing willingness to pay and optimal product configurations. If you need to prioritise a list, use MaxDiff. If you need to model product or pricing decisions, use conjoint.
What Is MaxDiff Analysis?
MaxDiff (Maximum Difference Scaling, sometimes called best-worst scaling) is a survey technique that asks respondents to identify the most and least important items from a small set shown simultaneously. This is repeated across multiple sets until every item in your list has been evaluated enough times to produce a reliable importance score.
How MaxDiff Works
A respondent sees a screen showing four to six items — product features, marketing messages, service attributes, or any list of things you want to rank. They select the one they find most important and the one they find least important. The survey cycles through different combinations of the same item pool until the forced trade-offs reveal a stable ranking.
The output is a set of importance scores (sometimes expressed as zero-centred utilities or rescaled to sum to 100) that show which items are clearly valued most, which cluster in the middle, and which sit at the bottom.
What MaxDiff Is Good At
- Feature prioritisation. Understanding which product features customers value most before committing engineering resources.
- Message testing. Ranking marketing messages or value propositions by resonance with a target audience.
- Attribute importance as a precursor to conjoint. Identifying the two to four attributes that matter enough to include in a more complex conjoint study.
- Comparing preferences across segments. MaxDiff scores are highly comparable across demographic groups, making segment analysis straightforward.
MaxDiff produces clean, interpretable data because the forced choice format eliminates the acquiescence bias that plagues rating scales (where respondents tend to rate everything "important").
What Is Conjoint Analysis?
Conjoint analysis is a method for measuring how people make trade-off decisions when evaluating full product or service descriptions. Instead of evaluating individual attributes in isolation, respondents evaluate complete profiles — combinations of attributes and levels — and their choices reveal the relative value they place on each dimension.
How Conjoint Analysis Works
The most common form is Choice-Based Conjoint (CBC). Respondents see sets of two to four product profiles simultaneously and choose their preferred option. Each profile is defined by a specific combination of attribute levels (for example: price = $49/month, storage = 50 GB, support = email only). By varying which levels appear together across many choice tasks, the analysis isolates the contribution of each attribute level to overall preference — these are called part-worth utilities.
From part-worth utilities, researchers can:
- Calculate each attribute's relative importance
- Simulate how different product configurations would perform against each other
- Estimate price sensitivity and willingness to pay
- Run market share simulations across competitive scenarios
What Conjoint Analysis Is Good At
- Pricing sensitivity. Understanding what customers will pay for different feature combinations, and where price resistance begins.
- Product configuration. Designing the feature bundle that maximises preference in a specific market segment.
- Packaging decisions. Determining how to tier a product offering across plan levels.
- Competitive positioning. Modelling how a new product configuration would perform against existing market alternatives.
MaxDiff vs Conjoint: Comparison Table
| Dimension | MaxDiff | Conjoint Analysis |
|---|---|---|
| Question answered | Which items are most / least important? | How do people trade off attributes in product decisions? |
| Task format | Select best and worst from a small item set | Choose among full product profiles |
| Output | Importance scores / rankings for each item | Part-worth utilities, attribute importance, market simulations |
| Best for | Feature prioritisation, message testing, attribute screening | Pricing research, product design, packaging decisions |
| Design complexity | Low to moderate | Moderate to high |
| Sample size needed | 150–300 per segment for stable rankings | 250–500+ per segment; larger for adaptive designs |
| Cost and time | Lower — simpler design and analysis | Higher — requires experimental design and utility computation |
When to Use MaxDiff
Use MaxDiff when your core question is: "Which of these things matters most to our audience?"
MaxDiff is the right method when:
- You have a list of 8 to 30 items you need to rank by importance or preference
- Your stakeholders need a clear priority order, not a trade-off model
- You are screening attributes before a conjoint study and want to drop the ones that do not drive decisions
- You are testing marketing messages and need to know which resonate most before you commit to a campaign
- Your timeline or budget does not support a full conjoint design
The method is particularly robust for comparing importance across customer segments. Because MaxDiff forces explicit trade-offs, the scores reflect genuine preference rather than courtesy ratings.
When to Use Conjoint Analysis
Use conjoint analysis when your core question is: "How will customers choose between different product configurations, and what are they willing to pay?"
Conjoint is the right method when:
- You are making a pricing or product packaging decision that requires modelling trade-offs
- You need to understand not just what customers want, but what they will sacrifice to get it
- You are designing or redesigning a product and need to know which feature combinations maximise preference
- You need a market simulation tool — the ability to test "what if" scenarios against a defined competitive set
- Your research will inform a high-stakes investment in product development or go-to-market strategy
Conjoint produces richer output than MaxDiff, but it requires a larger sample, a more careful experimental design, and more analytical expertise to interpret correctly.
When to Use Both Together
The most rigorous research programmes use MaxDiff and conjoint sequentially.
Phase 1 — MaxDiff for attribute screening. If your product has ten or more potential attributes, running conjoint on all of them simultaneously produces overwhelming choice tasks and unreliable data. MaxDiff first narrows the field. You run MaxDiff on the full attribute list and identify the four to six attributes that actually drive decisions. These become the inputs to your conjoint design.
Phase 2 — Conjoint for trade-off modelling. With a focused attribute set validated by MaxDiff, your conjoint study is more efficient, less fatiguing for respondents, and produces cleaner utility estimates.
This two-phase approach is standard practice in product development research, pricing studies, and segmentation projects where the attribute landscape is broad and the decision stakes are high.
Practical Considerations
Sample Size
MaxDiff is more forgiving on sample size. A study with 150 to 200 respondents can produce reliable importance rankings for most item sets. Conjoint analysis generally requires 250 to 500 respondents per analysis segment to generate stable part-worth utilities, and adaptive conjoint designs may require larger samples still.
If your research budget is constrained, MaxDiff may be the only viable option. If you are making a significant product or pricing decision, investing in the sample size required for conjoint is usually worthwhile.
Survey Design Complexity
MaxDiff surveys can be designed by a researcher with a working knowledge of survey methodology. The key decisions — item pool size, set size, number of sets per respondent — follow well-documented guidelines.
Conjoint designs require more expertise. Attribute selection, level choice, orthogonal or D-optimal design, prohibition constraints, and the choice between CBC, full-profile, and adaptive formats all affect data quality significantly. Errors at the design stage cannot be corrected after fieldwork.
Respondent Experience
MaxDiff tasks feel intuitive. Respondents understand "pick your top and bottom choice from this list" immediately. Conjoint tasks — choosing between full product profiles with multiple varying attributes — are more cognitively demanding. Studies with more than twelve to fifteen choice tasks per respondent see elevated abandonment rates and lower data quality in later tasks.
How onlinesurvey.ai Supports Both Methods
Running MaxDiff or conjoint studies requires more than a basic survey builder. Both methods depend on:
- Rotation and randomisation logic to ensure each item or profile appears the right number of times and in varying positions
- Quota controls to manage sample composition across segments
- Branching and skip logic to embed MaxDiff or conjoint blocks within a broader survey without exposing screener questions to unqualified respondents
- Export formats compatible with analysis tools — including structured data export for part-worth utility computation
onlinesurvey.ai's advanced survey logic supports the experimental design structures that MaxDiff and conjoint require: randomised item set rotation, within-survey quota management, and conditional branching that keeps the respondent experience clean regardless of study complexity. For teams running enterprise research programmes, the platform's unlimited surveys and response capacity on the Pro plan mean you can run screening and main study phases without managing separate accounts or hitting arbitrary caps.
Once your conjoint or MaxDiff data is collected, onlinesurvey.ai's AI analysis layer turns response data into a narrative executive summary — identifying key findings, patterns, and segment-level differences without requiring manual tabulation.
Frequently Asked Questions
What is the main difference between MaxDiff and conjoint analysis?
MaxDiff measures importance: it tells you which items from a list matter most to respondents by forcing best-and-worst choices. Conjoint analysis measures trade-off preferences: it tells you how respondents choose between full product or service descriptions and what they are willing to sacrifice to get what they want most. MaxDiff answers "what is important?"; conjoint answers "how do people make real decisions between options?"
Can I run a MaxDiff study with a standard survey tool?
You can approximate MaxDiff with a standard survey tool, but purpose-built MaxDiff requires randomised rotation logic that presents each item the right number of times in balanced sets. Without that logic, your importance scores will be biased by position effects and uneven exposure. A survey platform with advanced logic controls — randomisation, set rotation, and item balancing — is required for a methodologically sound MaxDiff study.
How many attributes can I include in a conjoint study?
Most conjoint studies use four to eight attributes. Beyond eight, choice tasks become too cognitively demanding and data quality deteriorates. If your product has more than eight candidate attributes, use MaxDiff first to screen down to the attributes that most strongly drive decisions, then design your conjoint study around the top four to six. This two-stage approach produces better data and a better respondent experience.
What sample size do I need for MaxDiff vs conjoint?
For MaxDiff, 150 to 200 respondents per analysis segment is typically sufficient for stable importance rankings. For conjoint, plan for 250 to 500 per segment, depending on the number of attributes and levels — more complex designs need larger samples. If you are planning to run market simulations or segment-level analysis, size up accordingly. Your conjoint software provider or methodologist should give you an explicit sample size recommendation based on your design.
When does it make sense to run both methods in the same research programme?
The most common scenario is when you are conducting large-scale product or pricing research with a broad attribute landscape. MaxDiff runs first to identify which four to six attributes from a longer list actually drive decisions. Those top attributes become the inputs to a conjoint design. This sequential approach prevents your conjoint study from being overwhelmed by attributes that respondents do not actually care about, and produces cleaner, more actionable trade-off data.
Is conjoint analysis only used for product research?
No. Conjoint analysis is used wherever people make trade-off decisions across defined options. Common applications outside product design include: pricing research for services and subscriptions, packaging and bundle design, healthcare treatment preference studies, employee benefits programme design, and communications testing where messages are combined with context variables. The core logic — exposing full profiles and measuring choice — applies wherever there are multiple attributes that jointly influence a decision.