What does IPMA tell you that path coefficients do not?

Path coefficients tell you importance only. IPMA adds performance, the current standing of each construct, so you can prioritize improvements rather than just rank effects. A construct can be very important yet already perform well, which changes where you should act.

What is importance in IPMA?

The total effect of a construct on the chosen target, including both direct and indirect paths through mediators. It is the unstandardized total effect.

How is performance calculated?

It is the average latent variable score rescaled to a 0 to 100 range, where 0 and 100 correspond to the lowest and highest possible values on the measurement scale. This requires consistently coded indicators.

PLS-SEM METHOD

IPMA: importance-performance map analysis

Importance-performance map analysis (IPMA) extends PLS-SEM results by combining how important each construct is for a chosen target with how well it currently performs. It turns path coefficients into a managerial priority map. This page explains the two axes, how to read the map, the indicator level, and the common pitfalls.

What IPMA adds to a PLS-SEM model

A standard PLS-SEM model tells you how strongly each construct drives a target, through total effects. IPMA (Ringle & Sarstedt 2016) adds a second dimension: performance, the average standing of each construct rescaled to a 0 to 100 scale. Plotting importance on the x-axis against performance on the y-axis produces a map that answers a practical question: where should you act first?

Constructs that are highly important but perform poorly are the priorities: improving them yields the biggest expected gain in the target. Constructs that already perform well, or that matter little, are lower priorities.

A worked example

For the target construct loyalty, satisfaction has high importance (a total effect of 0.46) but a performance of only 62 out of 100. It sits in the high-importance, low-performance quadrant, which makes it the first place to act.

Trust, by contrast, already performs at 84 with lower importance, so it is a keep-up-the-good-work construct rather than a priority.

Satisfaction (high importance, performance 62) sits in the priority quadrant; trust (performance 84) is already strong.

How to read the map

Each construct is a point: importance (its total effect on the target) on the x-axis, performance (0 to 100) on the y-axis. The quadrants guide where to act.

High importance, low performance: The top priority. These constructs move the target the most and have the most room to improve. Act here first.
High importance, high performance: Keep up the good work. Important and already strong, so protect them.
Low importance, low performance: Low priority. Weak drivers of the target, so improving them yields little.
Low importance, high performance: Possible over-investment. Strong but not very important, so resources might be better spent elsewhere.

Importance is the total effect, not just the direct path, so it includes effects routed through mediators. Always read the map relative to the one target construct you chose.

The two levels

Construct level

Importance (total effect on the target) and performance (rescaled mean score) for each construct. The headline map.

Indicator level

The same logic applied to individual indicators, using their rescaled performance and importance. Useful for pinpointing which specific item to improve.

Importance

The unstandardized total effect of a construct on the target. It captures direct plus indirect paths.

Performance

The construct’s average latent score rescaled to 0 to 100, where 100 is the best possible value on the measurement scale.

Running IPMA step by step

Estimate your PLS-SEM model and choose the target construct whose drivers you want to prioritize.
Make sure all indicators are coded in the same direction and on a known scale, so performance can be rescaled meaningfully.
Run IPMA to get importance and performance for every construct that affects the target.
Read the map and focus on constructs in the high-importance, low-performance quadrant.
Drill down to the indicator level to see which specific items drive the priority constructs.

Common pitfalls

Running IPMA when indicators are reverse-coded or on inconsistent scales, which makes the performance rescaling meaningless.
Reading importance as the direct effect only. IPMA uses total effects, which include indirect paths.
Forgetting that the map is relative to one target construct; a different target gives a different map.
Acting on performance alone, or on importance alone. The point of IPMA is to combine the two.

IPMA in OpenPLS

OpenPLS includes an IPMA panel for any target endogenous construct. It gives importance (total effect) and performance (rescaled to 0 to 100) at both the construct and the indicator level, with the common scale bounds applied. Free and runs in the browser.

Build an importance-performance map for your model

Estimate your model in the visual editor, pick a target, and get the IPMA map at construct and indicator level in one run. Free, no install.

Open OpenPLS What is PLS-SEM?

Frequently asked questions

What does IPMA tell you that path coefficients do not?: Path coefficients tell you importance only. IPMA adds performance, the current standing of each construct, so you can prioritize improvements rather than just rank effects. A construct can be very important yet already perform well, which changes where you should act.
What is importance in IPMA?: The total effect of a construct on the chosen target, including both direct and indirect paths through mediators. It is the unstandardized total effect.
How is performance calculated?: It is the average latent variable score rescaled to a 0 to 100 range, where 0 and 100 correspond to the lowest and highest possible values on the measurement scale. This requires consistently coded indicators.