Sign In

AI Readiness Guide · Appendix D

AI Use-Case Prioritization Worksheet.

Compare AI opportunities using organizational value, readiness, feasibility, measurability, risk, cost, and strategic alignment before choosing what deserves to become a pilot.

Developed by FridAI Learning Center Published Full score 65 points Quick screen 25 points
On this page

Once an organization begins looking for artificial intelligence opportunities, ideas can accumulate quickly.

Sales may have ideas. Operations may have ideas. Human resources, finance, employees, leadership, and vendors may all identify additional applications.

Before long, the organization may have 10, 20, or 50 possible AI use cases.

That is not a problem. The problem is trying to pursue all of them.

AI readiness includes the ability to prioritize.

This worksheet compares potential AI opportunities using:

  • Organizational value.
  • Frequency.
  • Process readiness.
  • Data readiness.
  • Technical feasibility.
  • Workforce readiness.
  • Measurability.
  • Speed to learning.
  • Reversibility.
  • Risk.
  • Strategic alignment.
  • Internal capability.
  • Cost and resource feasibility.

The purpose is not to create a mathematically perfect ranking. The purpose is to force better discussion.

Part 1

Start With the Problem

Do not begin by naming the AI technology. Begin by describing the problem.

Part 2

Confirm That AI Is Actually Appropriate

Could the problem be solved more simply?

Possible non-AI solutions might include:

  • Process redesign.
  • Better training.
  • Standard software.
  • System integration.
  • Workflow automation.
  • A spreadsheet.
  • A database.
  • Eliminating unnecessary work.

If a simpler solution appears likely to solve the problem effectively, evaluate that option first.

Part 3

Scoring Method

Rate each category from 1 to 5.

1

Very Weak

The use case performs poorly in this area.

2

Weak

Significant barriers exist.

3

Moderate

The use case is workable but has meaningful limitations.

4

Strong

The use case performs well in this area.

5

Very Strong

The use case is especially favorable in this area.

Risk uses the opposite direction

A risk score of 1 means very high risk. A risk score of 5 means very low risk.

A higher overall score therefore indicates a more attractive early opportunity.

Part 4

Organizational Value

Question

If this works, how meaningful will the improvement be?

Consider employee time saved, cost reduction, revenue, customer experience, employee experience, quality, capacity, risk reduction, decision quality, and strategic advantage.

  1. 1
    Minimal value

    Little organizational impact.

  2. 2
    Limited value

    Some improvement, but relatively minor.

  3. 3
    Moderate value

    Noticeable operational benefit.

  4. 4
    High value

    Meaningful organizational improvement.

  5. 5
    Very high value

    Substantial operational, financial, strategic, or stakeholder impact.

/5

Part 5

Frequency

Question

How often does the problem or task occur?

Small improvements can become significant when they are repeated often.

  1. 1
    Very rare

    Occurs only occasionally.

  2. 2
    Infrequent

    Occurs several times per year.

  3. 3
    Regular

    Occurs monthly or weekly.

  4. 4
    Frequent

    Occurs several times per week or daily.

  5. 5
    Very frequent or high volume

    Occurs repeatedly throughout normal operations.

/5

Part 6

Process Readiness

Question

Do we understand the current workflow well enough to improve it?

Consider process documentation, standardization, known bottlenecks, decision points, common exceptions, and employee understanding.

  1. 1
    Poorly understood

    The process is inconsistent or unclear.

  2. 2
    Limited understanding

    Major questions remain.

  3. 3
    Moderate

    The process is generally understood but contains variability.

  4. 4
    Strong

    The process is well understood and reasonably standardized.

  5. 5
    Very strong

    The process is clear, documented, measurable, and repeatable.

/5

Part 7

Data Readiness

Question

Is the required information available and usable?

Consider access, quality, completeness, consistency, historical depth, sensitivity, and ownership.

  1. 1
    Data unavailable

    Required information does not exist or cannot be accessed.

  2. 2
    Significant data problems

    Major preparation is required.

  3. 3
    Usable with caution

    Enough data may exist for a limited pilot.

  4. 4
    Good

    Data is reasonably accessible and reliable.

  5. 5
    Strong

    Data is understood, reliable, accessible, and appropriately governed.

/5

Part 8

Technical Feasibility

Question

How practical is the technology implementation?

Consider existing tools, infrastructure, integration, connectivity, vendor availability, internal expertise, and maintenance requirements.

  1. 1
    Very difficult

    Requires substantial infrastructure or specialized capability.

  2. 2
    Difficult

    Major technical challenges exist.

  3. 3
    Moderate

    Implementation is possible with manageable support.

  4. 4
    Straightforward

    Most required technology already exists or is readily available.

  5. 5
    Very straightforward

    The use case can be tested quickly with accessible technology.

/5

Part 9

Workforce Readiness

Question

Are the people who will use or be affected by the solution reasonably prepared?

Consider employee interest, manager support, training needs, workflow fit, trust, and potential resistance.

  1. 1
    Very low readiness

    Significant workforce barriers exist.

  2. 2
    Low readiness

    Substantial preparation is needed.

  3. 3
    Moderate

    Employees can likely participate with training and support.

  4. 4
    Strong

    Employees are reasonably prepared and supportive.

  5. 5
    Very strong

    Employees understand the problem, support experimentation, and can adopt quickly.

/5

Part 10

Measurability

Question

Can we determine whether the AI use case actually worked?

Consider whether the organization can establish a baseline, clear success criteria, and meaningful outcome measures.

  1. 1
    Very difficult to measure

    Success would be largely subjective.

  2. 2
    Difficult

    Only indirect evidence is available.

  3. 3
    Moderate

    Some useful measures exist.

  4. 4
    Strong

    Clear quantitative or qualitative measures are available.

  5. 5
    Very strong

    Baseline and outcome measures are clear, credible, and easy to compare.

/5

Part 11

Speed to Learning

Question

How quickly can the organization obtain useful evidence?

A pilot that produces feedback rapidly may be more useful as an early project.

  1. 1
    Very slow

    Meaningful results may require a year or more.

  2. 2
    Slow

    Several months may be required.

  3. 3
    Moderate

    Evidence could emerge within a few months.

  4. 4
    Fast

    Feedback could emerge within several weeks.

  5. 5
    Very fast

    Frequent use or transactions allow rapid learning.

/5

Part 12

Reversibility

Question

If the AI produces an incorrect result, can the organization recover easily?

  1. 1
    Difficult to reverse

    Errors may create lasting consequences.

  2. 2
    Limited reversibility

    Correction is possible but costly or difficult.

  3. 3
    Moderate

    Most mistakes can be corrected.

  4. 4
    Highly reversible

    Errors can usually be corrected quickly.

  5. 5
    Very highly reversible

    Drafts or recommendations can be rejected with little consequence.

/5

Part 13

Risk

Question

What is the potential consequence of failure, misuse, or error?

Consider privacy, security, legal exposure, financial loss, safety, fairness, reputation, employee impact, customer impact, and regulatory consequences.

Higher score = lower risk
  1. 1
    Very high risk

    Errors could cause serious legal, financial, safety, privacy, or human consequences.

  2. 2
    High risk

    Meaningful consequences require strong controls.

  3. 3
    Moderate risk

    Risks are manageable with governance and human review.

  4. 4
    Low risk

    Consequences of error are generally limited.

  5. 5
    Very low risk

    Errors are detectable, reversible, and unlikely to create meaningful harm.

/5

Part 14

Strategic Alignment

Question

Does the use case support something the organization already considers important?

Consider alignment with growth, productivity, customer service, employee retention, quality, innovation, cost control, capacity, risk management, or mission impact.

  1. 1
    Little alignment

    Disconnected from organizational priorities.

  2. 2
    Limited alignment

    The connection is indirect.

  3. 3
    Moderate alignment

    Supports a meaningful but secondary priority.

  4. 4
    Strong alignment

    Directly supports an important priority.

  5. 5
    Very strong alignment

    Directly advances one of the organization's highest strategic priorities.

/5

Part 15

Internal Capability Fit

Question

Does the organization have enough internal capability to participate effectively in the project?

Consider ownership, subject-matter expertise, data literacy, technical knowledge, AI champions, and the ability to evaluate results.

  1. 1
    Major capability gap

    The organization would depend almost entirely on outside expertise.

  2. 2
    Significant gap

    Important skills or ownership are missing.

  3. 3
    Moderate

    Outside support may be needed, but internal participation is realistic.

  4. 4
    Strong

    Most capabilities required for a pilot exist.

  5. 5
    Very strong

    Strong ownership and expertise already exist.

/5

Part 16

Cost and Resource Feasibility

Question

Can the organization realistically afford the pilot and potential next steps?

Consider software, vendor support, consulting, employee time, training, data preparation, integration, and ongoing support.

  1. 1
    Very difficult

    Cost or resource requirements are likely prohibitive.

  2. 2
    Difficult

    Significant investment is required.

  3. 3
    Manageable

    The organization could fund a controlled pilot.

  4. 4
    Strong

    Investment is reasonable relative to expected value.

  5. 5
    Very strong

    The use case can be tested inexpensively with favorable potential economics.

/5

Part 17

Total Score

The scores below populate automatically from Parts 4–16.

Criterion Score
Organizational Value /5
Frequency /5
Process Readiness /5
Data Readiness /5
Technical Feasibility /5
Workforce Readiness /5
Measurability /5
Speed to Learning /5
Reversibility /5
Risk /5
Strategic Alignment /5
Internal Capability Fit /5
Cost and Resource Feasibility /5
Maximum Score 65

/65

%

Score ÷ 65 × 100 = Percentage

Part 18

Interpreting the Score

The score should guide discussion rather than make the decision automatically.

80–100%

Strong Candidate

The use case appears to combine meaningful value, reasonable readiness, manageable risk, and strong feasibility.

It may be a strong candidate for near-term pilot development.

65–79%

Promising Candidate

The opportunity appears worthwhile, but one or more readiness issues should be reviewed.

Determine whether the gaps should be addressed before or during the pilot.

50–64%

Develop Further

The use case may have value, but important feasibility, readiness, risk, or measurement questions remain.

Consider improving the use case before investing significantly.

Below 50%

Low Near-Term Priority

The use case may be too risky, difficult, poorly defined, or insufficiently valuable for immediate pursuit.

Keep it in the opportunity backlog if longer-term potential exists.

Part 19

Do Not Let the Score Override Judgment

A use case with a high score may still be inappropriate.

  • A high-risk legal issue may require specialized review.
  • Regulatory restrictions may prohibit the approach.
  • An important ethical concern may exist.

A lower-scoring project may also deserve attention because it addresses a strategically critical issue.

The score is decision support. Not decision replacement.

Part 20

Compare Multiple Use Cases

Use Case Score / 65 Value Feasibility Risk Strategic Alignment Decision

Part 21

Plot the Use Cases

Place each opportunity on a matrix using feasibility on the horizontal axis and organizational value on the vertical axis.

Organizational Value · Low → High
High Value / Low Feasibility

Strategic Opportunities

Important opportunities that require readiness improvement before implementation.

High Value / High Feasibility

Priority Opportunities

Usually the strongest candidates for near-term pilots.

Low Value / Low Feasibility

Low Priority

Usually poor candidates for near-term investment.

Low Value / High Feasibility

Easy Wins

Useful for learning, but should not consume disproportionate resources.

Feasibility · Low → High

Part 22

Add Risk to the Conversation

High Value + High Feasibility + Low Risk

This combination may be an excellent starting point.

High Value + High Feasibility + High Risk

The opportunity may still matter, but should involve stronger governance or a narrower pilot.

Low Value + High Risk

This combination should generally receive very low priority.

Risk can change the decision even when value and feasibility appear attractive.

Part 23

Identify the Strongest First Pilot

The strongest first pilot may not be the most exciting use case. It is usually the one that creates meaningful learning with reasonable risk.

Part 24

First-Pilot Decision

Part 25

Define the Pilot Question

The pilot should answer one primary question.

Can improve from to without creating unacceptable ?

Example

Can AI-assisted reporting reduce average preparation time from four hours to two hours without increasing material errors?

A clear question creates a clear evaluation.

Part 26

Define Success

Part 27

Define the Smallest Useful Pilot

What is the smallest experiment that would tell us whether this idea deserves more investment?

Avoid building a large implementation before answering the core question.

Part 28

Identify Readiness Actions Before Launch

Action 1

Action 2

Action 3

Actions may include cleaning a dataset, training managers, approving a tool, creating AI-use guidance, documenting the process, establishing a baseline, or completing a vendor review.

Part 29

Use-Case Decision Gate

Before approving a pilot, confirm:

If several boxes remain unchecked, prepare further before launching.

Part 30

Opportunity Backlog

Not every worthwhile use case should begin immediately.

Opportunity Priority Reason for Waiting What Must Change? Review Date

This prevents good ideas from being forgotten while protecting the organization from doing too much at once.

Part 31

Reevaluate the Backlog Periodically

A use case that is unattractive today may become compelling later.

  • Data improves.
  • Technology costs fall.
  • A new vendor emerges.
  • Employees gain capability.
  • A related system is implemented.
  • The organizational priority becomes more important.
  • A regulatory issue becomes clearer.

Revisit the opportunity backlog periodically. AI prioritization is not a one-time decision.

Part 32

Avoid Common Prioritization Mistakes

Choosing the Flashiest Idea

An impressive demonstration does not guarantee organizational value.

Starting With the Most Difficult Problem

Transformation is not required for the first pilot.

Selecting Only Easy Projects

A project can be so trivial that no one cares about the result.

Ignoring Data Until Later

A great idea without usable data may not be ready.

Ignoring Employees

A technically strong solution may fail if the workflow does not work for people.

Ignoring Risk

High value does not erase high consequence.

Ignoring Measurement

If success cannot be defined, evaluation becomes difficult.

Pursuing Too Many Projects

Organizational learning capacity is limited.

Assuming AI Is Required

A simpler solution may be better.

Letting a Score Make the Decision

Structured judgment is still judgment.

Part 33

A Fast Five-Minute Screening Tool

Organizations with many ideas can use this preliminary screen before completing the full worksheet.

Value

Would success matter?

Feasibility

Can we realistically test it?

Data

Do we have usable information?

Measurability

Can we determine whether it worked?

Risk

How manageable is the error?

Quick Score

Five preliminary criteria

Use the quick screen to identify which ideas deserve deeper evaluation.

Part 34

A Simple Leadership Question

Which project would teach us the most while putting the least important things at risk?

That question captures the purpose of early AI adoption.

The first projects should create both value and learning.

Part 35

Final Use-Case Summary

/65
Recommended decision

The Purpose of Prioritization

Artificial intelligence creates more possibilities than most organizations can realistically pursue.

As AI capabilities expand, the opportunity list will become longer, not shorter.

The organizations that succeed will not necessarily be the ones that pursue the most ideas. They will be the ones that become good at choosing.

  • Choosing problems that matter.
  • Choosing applications that are feasible.
  • Choosing risks they understand.
  • Choosing experiments that produce evidence.
  • Choosing when to invest.
  • Choosing when to wait.
  • Choosing when to say no.

That discipline protects the organization from both extremes: doing nothing because AI feels overwhelming and doing everything because AI feels exciting.

The goal is to direct organizational attention toward the opportunities most likely to create meaningful value.

Find the problem Evaluate the opportunity Prioritize deliberately Pilot what deserves testing Let evidence decide

Find the problem. Evaluate the opportunity. Prioritize deliberately. Pilot what deserves to be tested. Then let evidence determine what comes next.

Last reviewed:

This worksheet supports internal prioritization and does not replace specialized legal, privacy, security, financial, or technical review.

← Previous appendix Appendix C: Data Readiness Checklist and Inventory Worksheet

Identify, evaluate, classify, and prepare the data required for a responsible AI use case.

Continue to Appendix E → 90-Day AI Readiness Action Plan

Turn readiness findings into accountable actions, owners, milestones, and measurable progress.