MaxDiff Feature Analysis
Identify your product's 'Must-Haves' vs 'Don't Cares'. Best-Worst scaling forces trade-offs to reveal true preference order.
Feature Counts
Preference Utility Scores
Ranking Priorities
Positive = Value Driver. Negative = Value Detractor/Indifference.
The Critical Few
The top 2-3 features usually carry 80% of the decision weight. Focus your marketing here.
The Trivial Many
Features clustered around 0 are interchangeable. Don't base your differentiation strategy on them.
The Cost Cutters
Features with deep negative scores are candidates for removal to save margin.
Execution Steps
Enter the list of features or attributes you are testing.
Input the raw counts of how many times each feature was chosen as 'Most Important' (Best) and 'Least Important' (Worst) in your survey.
The chart calculates a Net Utility Score (Best - Worst).
Features with high positive bars are critical drivers. Features with negative bars are potential areas to cut costs.
Pro Strategy
- If 'Price' is an attribute, it usually dominates. Consider running price separately in a Van Westendorp or Conjoint study.
- Features near zero are 'Nice to Haves' but not deal-breakers.
- Stop investing in features with highly negative scores. Your customers actively dislike them or consider them waste.
Core Concepts
Best-Worst Scaling
A method where respondents choose the best and worst option from a subset. It is far more discriminatory than rating scales (1-5 stars).
Utility Score
A dimensionless measure of preference. The distance between scores indicates the magnitude of preference difference.
Turf Analysis
Often paired with MaxDiff to find the subset of features that reaches the widest unique audience (Total Unduplicated Reach and Frequency).
What is MaxDiff Feature Analysis?
Maximum Difference Scaling (MaxDiff) is a discrete choice model. It assumes that when a respondent looks at a set of items, they evaluate all possible pairs and choose the pair with the maximum difference in utility (the best and the worst). This provides a ratio-scaled ranking of items.
Best For
- • Prioritizing a product roadmap.
- • Determining which claims to put on packaging.
- • Identifying features to cut for a 'Lite' version.
Limitations
- • Doesn't tell you *how much* they like the best item, just that it is #1.
- • Doesn't measure willingness to pay directly.
- • Can be tedious for respondents if the list is too long.
Alternative Methods
Kano Model
Classifies features into Basic, Performance, and Delighters based on satisfaction.
Conjoint Analysis
Better if you need to trade off features against price specifically.
Industry Applications
See how this methodology generates real revenue uplift in different sectors.
Smartphone Feature Prioritization
Engineering wanted to upgrade the screen resolution, marketing wanted better battery life.
MaxDiff showed 'Battery Life' had 3x the utility score of '4K Screen' for the target audience.
Hotel Amenities
Hotel chain wanted to cut costs. Options: Cut free breakfast, gym, or pool.
Analysis revealed 'Free Breakfast' was a top 3 driver, while 'Pool' was bottom 3 (most people don't use it).