What Is Importance-Performance Analysis?
Importance-performance analysis (IPA) is a strategic framework that plots attributes on a two-dimensional grid based on how important they are to customers and how well an organization performs on each one. Developed by Martilla and James in 1977, it remains one of the most practical tools in customer research because it translates complex satisfaction data into a visual prioritization map. Each attribute lands in one of four quadrants, and each quadrant carries a clear strategic implication: concentrate here, keep up the good work, low priority, or possible overkill. The power of IPA is its ability to turn dozens of attribute ratings into a single actionable visual that non-researchers can understand immediately.
Why Importance-Performance Analysis Matters
Satisfaction research often produces overwhelming volumes of data, ratings on 20+ attributes across multiple segments. Executives don't need all 20 ratings; they need to know where to invest. IPA compresses the data into a decision framework. It answers the question every stakeholder asks: "What should we fix first?" And it does so visually, which makes the answer stick in a way that tables and ranked lists don't.
How Importance-Performance Analysis Works
Building the Grid
The IPA grid has two axes:
- X-axis: Importance: how important each attribute is to customers. This can be measured directly (stated importance from a survey question) or derived (from correlation or regression analysis against overall satisfaction).
- Y-axis: Performance: how well the organization performs on each attribute, typically the mean satisfaction or quality rating.
Each attribute is plotted as a point on the grid. The grid is divided into four quadrants by drawing crosshairs at the midpoint, grand mean, or another meaningful threshold.
The Four Quadrants
Quadrant I: Concentrate Here (high importance, low performance) These attributes matter most to customers, and you're underdelivering. They represent the highest-priority improvement opportunities. Every unit of investment here produces the biggest return in customer satisfaction.
Quadrant II: Keep Up the Good Work (high importance, high performance) These are your strengths, attributes that drive satisfaction and where you're performing well. The strategic imperative is maintenance, not complacency. Letting performance slip on these attributes is costly because they're the ones customers care about most.
Quadrant III: Low Priority (low importance, low performance) Poor performance here is less damaging because customers don't weight these attributes heavily. Improving them won't move the satisfaction needle much. Allocate resources elsewhere unless these attributes are trending toward higher importance.
Quadrant IV: Possible Overkill (low importance, high performance) You're performing well on things customers don't prioritize. This isn't necessarily bad, some of these attributes may be table stakes that don't differentiate but can't be dropped. But if resources are constrained, this quadrant reveals where you might be over-investing.
Stated vs. Derived Importance
The original IPA framework uses stated importance, directly asking customers to rate how important each attribute is. This works but has a known weakness: respondents tend to rate everything as highly important, compressing the importance axis and pushing most attributes into the top half of the grid.
Derived importance, calculating importance from the correlation between attribute performance and overall satisfaction, produces better differentiation. It reflects what actually drives satisfaction rather than what respondents claim matters. Many practitioners now use derived importance on the x-axis while keeping stated performance on the y-axis.
Interpreting the Grid
A few analytical considerations improve interpretation:
Where you draw the crosshairs matters. Using the scale midpoint (e.g., 5.0 on a 1-10 scale) as the divider often puts everything in the high-high quadrant because most ratings skew positive. Using the grand mean of all attributes creates a more balanced distribution and better differentiates relative priorities.
Attributes near the crosshairs are ambiguous. An attribute sitting right on the border between "Concentrate Here" and "Keep Up the Good Work" shouldn't be treated as definitively in either quadrant. Apply confidence intervals or sensitivity analysis for borderline cases.
Segment-level IPA reveals different priorities. Running IPA for your total sample and then separately for key segments (new vs. Loyal customers, different demographics) often shows that the priority quadrant map shifts significantly. What new customers need you to concentrate on may be different from what long-term customers value.
Dynamic IPA
For tracking studies, plot IPA results across waves to see how attributes move over time. An attribute that migrates from "Concentrate Here" to "Keep Up the Good Work" confirms that improvement initiatives are working. One that drifts from "Keep Up the Good Work" to "Concentrate Here" signals a performance decline on something customers still value.
When to Use Importance-Performance Analysis
- Customer satisfaction program reporting: translating attribute-level data into a strategic prioritization visual for leadership.
- Service quality improvement planning: identifying which service dimensions need the most attention.
- Product development prioritization: determining which features to enhance, maintain, or de-prioritize.
- Competitive analysis: plotting your IPA alongside a competitor's to identify relative strengths and vulnerabilities.
- Employee experience surveys: mapping workplace attribute importance against organizational performance.
Common Mistakes to Avoid
- Using only stated importance: self-reported importance lacks discrimination. Supplement or replace it with derived importance from correlation or regression analysis for a more useful grid.
- Treating quadrant boundaries as absolute: attributes near the crosshairs can shift quadrants with small changes in data. Don't build strategy around borderline placements without checking sensitivity.
- Ignoring the "Possible Overkill" quadrant: teams naturally focus on the "Concentrate Here" quadrant and ignore Q4. But reallocating over-investment from Q4 to Q1 is often the most efficient path to improvement.
Quali-Fi Support
Quali-Fi's analytics dashboard generates importance-performance grids automatically from satisfaction surveys that include paired importance and performance ratings. The platform supports both stated and derived importance calculations and lets you filter the grid by segment, wave, or any survey variable, so you can see how priorities shift across different customer groups.
Frequently Asked Questions
How many attributes should I include in an IPA?
Between 10 and 25 works best. Fewer than 10 may miss important dimensions. More than 25 clutters the grid and makes it hard to read. If your survey has 30+ attributes, group them into broader categories for the IPA visual and provide attribute-level detail in a supporting table.
Can I use IPA for competitive benchmarking?
Yes. Plot your performance and competitor performance on the same grid using shared importance scores. This reveals where you outperform competitors on high-importance attributes (competitive advantages) and where they outperform you (competitive vulnerabilities).
What software do I need for IPA?
Any tool that creates scatter plots can generate an IPA grid. Excel, Google Sheets, Tableau, Power BI, or R. The key is formatting: label each point clearly, draw quadrant lines at your chosen threshold, and label the quadrants. Quali-Fi's dashboard generates IPA grids natively from survey data.
Related Topics
- Key Driver Analysis
- Gap Analysis (Research)
- Data Visualization for Research
- Segmentation Analysis
- Perceptual Mapping
- Data Collection Methods
Generate importance-performance grids from your survey data automatically. Start your free 14-day Quali-Fi trial, no credit card required.