How to Build an AI Use Case Agenda by Area: From Inventory to Prioritized Portfolio
By Pedro Sonda, dooop
To build an AI use case agenda by area, start from business objectives, create an inventory of opportunities in each area using the same format, evaluate all with the same criteria (value, feasibility, risk, and execution capacity), assemble a single portfolio balanced between quick wins and structural bets, and transform the portfolio into a roadmap with responsible parties and periodic review. The key point is to use common criteria: without them, each area prioritizes differently and resources become scattered.
Updated in October 2026.
Why a Common Agenda Matters
In a Gartner survey of 432 leaders, difficulty identifying appropriate use cases was the main obstacle cited by organizations with low AI maturity (37%) (Gartner, 2025). BCG estimates that about 70% of AI’s potential value lies in core functions such as sales, marketing, manufacturing, supply chain, and pricing (BCG, 2025).
Interpretation: many companies do not know where to apply AI, and those that do tend to start with support areas, which concentrate less value. A single agenda helps direct effort where the results are.
dooop Framework: The Agenda in 6 Steps
Step 1: Start from Business Objectives
List the three to five strategic objectives for the next two years. Each use case should connect to at least one of them.
Step 2: Create an Inventory by Area, Using the Same Format
Use a standard form for all areas:
| Field | Example Entry |
|---|---|
| Area and Process | Commercial, proposal preparation |
| Problem | Proposals take five days; 30% return with condition errors |
| Type of Opportunity | Automation, decision support, analysis, or new service |
| Linked Business Objective | Increase conversion rate |
| Required Data | Proposal history, price table, policies |
| Potential Owner | Commercial manager |
| Indicator | Preparation time; error rate; conversion |
Step 3: Define Common Criteria
At dooop, discovery prioritizes by value, feasibility, acceptable risk, and execution capacity (dooop). Other criteria may be used, provided they are the same for all areas and agreed upon with leadership before evaluation.
| Criterion | Question |
|---|---|
| Value | How much does it contribute to business objectives? |
| Feasibility | Do data, technology, and process allow it? |
| Risk | What are the risks to customers, people, data, and reputation? |
| Execution Capacity | Do we have people, time, and an owner to carry it forward? |
Step 4: Position on the Value × Feasibility Matrix
| High Feasibility | Low Feasibility | |
|---|---|---|
| High Value | Do now | Prepare (data, processes) and do later |
| Low Value | Do if cheap and quick | Discard |
High-risk cases only rise in priority with defined controls.
Step 5: Assemble a Balanced Portfolio
| Type | Role in Portfolio | Indicative Proportion |
|---|---|---|
| Quick Wins | Show results and generate learning | Several, across various areas |
| Structural Improvements | Redesign core processes | Few, in highest-value areas |
| Differentiation Bets | Create competitive advantage | One or two |
The balance among these types is discussed in AI as a Productivity Tool or Competitive Advantage?.
Step 6: Transform into a Roadmap with Review Rhythm
- Responsible parties for each use case.
- Sequence considering dependencies, such as data that must be organized first.
- Quarterly review in a decision forum: what progressed, what to stop, what to add.
Common Mistakes
- Each area with its own list and criteria.
- Portfolio without discards: an agenda that discards nothing has not prioritized.
- Only quick wins: generate enthusiasm but rarely change outcomes.
- No review: the agenda becomes a document and ages within months.
What to Do Now
- Distribute the standard form to areas and request three to five opportunities from each.
- Agree on criteria with the executive team before evaluating.
- Assemble the first portfolio and schedule the first quarterly review.
To build the inventory and prioritize with shared criteria, learn about dooop’s discovery at dooop.com.br/discovery-de-oportunidades-de-ia.
Frequently Asked Questions
How many use cases should the agenda have? It depends on execution capacity. For most medium-sized companies, a few cases running simultaneously, with a larger inventory waiting, is more effective than many cases in parallel.
Who should decide the agenda? A forum with the executive team and area representatives, supported by IT, data, and risk. Areas propose; the forum decides using common criteria.
Also Read
About dooop
dooop is a consultancy based in Caxias do Sul (RS) that prepares leaders and organizations to transform artificial intelligence into business capability through strategy, governance, and training.