How we rate use cases
Every use case in the directory gets an Impact rating and an Ease rating. Each is High, Medium or Low. This page shows how the words are decided, so you can check any rating against your own situation.
The ratings describe the use case in general. They are not your score. Your own self-assessed GAAPx Score is explained in how the score works.
What Impact is
Impact is how much a use case improves the work. We score it on five criteria, named Flow, Productivity, Quality, Control and Business outcome.
Each criterion gets a score from 1 to 5, with a one-line reason, and a weight. Weights are set for each use case and add up to 100. The weight says how much that criterion counts for this use case. A control test counts Control most. A vendor portal counts Control and Productivity most.
What Ease is
Ease is how hard the use case is to start. We score it on three criteria: Technology available, Team and data ready, and Controls in place. Each also gets a score from 1 to 5, a reason and a weight.
A high Ease means the tools exist and little of the work is on your side. A low Ease means most of the work, such as cleaning data or writing down who may do what, is yours.
How the 1 to 5 score and the thresholds work
- Score each criterion from 1 to 5.
- Multiply each score by its weight.
- Add the results and divide by the total of the weights. That is the weighted score, from 1 to 5.
- Read the label from the score.
| Weighted score | Label |
|---|---|
| Below 2.5 | Low |
| 2.5 to 3.5 | Medium |
| Above 3.5 | High |
Both ends of the Medium band count as Medium, so exactly 2.5 and exactly 3.5 are Medium.
Worked example: continuous control monitoring
Impact:
| Criterion | Weight | Score | Weight x score |
|---|---|---|---|
| Flow | 5 | 2 | 10 |
| Productivity | 20 | 3 | 60 |
| Quality | 20 | 4 | 80 |
| Control | 50 | 5 | 250 |
| Business outcome | 5 | 2 | 10 |
| Total | 100 | 410 |
410 divided by 100 is 4.1, which is above 3.5, so Impact is High.
Ease:
| Criterion | Weight | Score | Weight x score |
|---|---|---|---|
| Technology available | 35 | 3 | 105 |
| Team and data ready | 35 | 2 | 70 |
| Controls in place | 30 | 3 | 90 |
| Total | 100 | 265 |
265 divided by 100 is 2.65, which sits between 2.5 and 3.5, so Ease is Medium.
Reference: criteria
Weights are set for each use case and are shown as shares of 100. The example column is the continuous control monitoring example above.
| Rating | Criterion | What it asks | Example weight |
|---|---|---|---|
| Impact | Flow | Does work move, or sit blocked? | 5 |
| Impact | Productivity | How much comparable work is completed for resources used? | 20 |
| Impact | Quality | Is it correct the first time? | 20 |
| Impact | Control | Is it authorized, supported and traceable? | 50 |
| Impact | Business outcome | Does it improve cash, close, reporting or decisions? | 5 |
| Ease | Technology available | Do tools for this exist today, and do they connect to common systems? | 35 |
| Ease | Team and data ready | Is the data clean enough, and will people do the new work? | 35 |
| Ease | Controls in place | Are the rules and approvals the tool needs already written down? | 30 |
How Prep follows from Ease
Prep is the short label on a use case that tells you how much to do before you start. It has no score of its own. It is read straight from Ease:
| Ease | Prep |
|---|---|
| High | Light |
| Medium | Some |
| Low | A lot |
How the five process facts are decided
Each use case page lists the same five process facts, in the same order: process steps, cycle time, exception loops, first-time-right rate and cost per invoice. Each fact gets one of four labels. They describe expected effects, not measured results.
| Label | Meaning |
|---|---|
| Potential direct benefit | The use case acts on this fact directly. Expected effect, not a measured result. |
| Possible downstream benefit | The fact may improve as a knock-on effect. Expected effect, not a measured result. |
| Depends on workflow design | The result turns on how you set up the workflow. Set only by explicit editorial copy. |
| Not directly addressed | The use case does not directly change this fact. |
How the label is chosen:
- Each Impact criterion feeds two or fewer facts. Flow feeds cycle time and exception loops. Productivity feeds process steps and cost per invoice. Quality feeds exception loops and first-time-right rate. Control feeds first-time-right rate. Business outcome feeds cycle time and cost per invoice.
- For each fact, we add up weight x score over the criteria that feed it.
- Wording in the use case's own list of measures it moves can put a fact ahead of the others. For example, a measure that mentions "exceptions" points at exception loops. Facts with such a cue rank ahead of facts without one, and weight x score orders the rest.
- The top-ranked fact reads Potential direct benefit. The next two read Possible downstream benefit. Ties follow the fixed order above. The rest read Not directly addressed.
- Cost per invoice measures invoice work. Outside Spend, Get Paid and The Close it reads Not directly addressed.
- Some pages set labels by hand and give the reason next to each. On those pages, a fact with no hand-set label reads Not directly addressed. Only hand-set labels can read Depends on workflow design. The reason on the page explains the label.
A use case may help at most three facts by default. Two use cases with the same Impact rating can move different facts.
What "Year one, in hours" means
Where a use case has one, the page shows a short worked figure called "Year one, in hours". It uses hours and counts only.
- It shows an illustrative company, not yours. The example uses a 300-person software company.
- Every input marked as an assumption is a guess you can replace. Use your own control count, your own testing hours and your own volumes.
- The result is the input hours multiplied through. For continuous control monitoring: 20 controls, 6 hours each by hand, 2 hours each after, is 20 x (6 minus 2) = 80 hours a quarter, and 320 hours a year.
- The figure counts time only. It leaves out errors caught earlier and other benefits, which are real but harder to size.
The figures show how the arithmetic works. They are not a forecast for your team.
What this is not
- It is not a standard. The criteria and weights are ours, set so that readers can compare use cases. No audit body or regulator has set them.
- It is not a guarantee. A High Impact rating means the use case tends to change a lot. It does not promise a result in your company.
- It is not peer data. No rating compares you with other companies, and none uses another company's figures.
- It is not a verdict on your team. A Low Ease means more groundwork. It says nothing about how well you work today.
- It is not fixed. If a score, weight or label changes, the change is recorded in the changelog.
Your own starting point can move any rating. Treat each one as a place to begin the conversation.