The opportunity catalog has no prioritization built in. The Value × Feasibility matrix is the tool that turns a list into a decision — and positions the consultant as the one leading the diagnosis, not just describing it.
Illustrative diagram — Value × Feasibility matrix with the four decision quadrants.
💰 Value Axis: Impact, Frequency, Cost, and Revenue
Value isn't an impression — it's an estimate. Four dimensions make up the matrix's vertical axis: impact, frequency, avoided cost, and enabled revenue.
Impact on the end customer
Does the change improve quality, speed, or satisfaction for those who use the process’s output? Customer impact is the strongest argument for leadership.
Frequency × unit cost
Annual frequency × cost per occurrence = total cost of the pain point. This is the most concrete number in the business case. Example: 200 analyses/day × R$15 = R$3k/day = ~R$750k/year.
Cost avoided
How much is spent on rework, correcting errors, overtime, and outsourcing? This is the easiest figure to calculate using data from the process mapping.
Enabled revenue
What automation makes possible that isn’t possible today — the ability to serve more clients, launch products faster, or reduce delivery time?
💡 Qualitative value counts too
Not everything has an easy number. Reputational risk reduction, team satisfaction, regulatory compliance — record these as "high/medium/low qualitative value" when they can’t be quantified. But try to quantify them first.
🔧 Feasibility axis: data, technology, integration, change
Feasibility is the denominator of ROI. A high-value project with zero feasibility delivers nothing. Four dimensions make up the horizontal axis: data, technical complexity, integration, and organizational change.
Data readiness
Does the data exist? Is it in an accessible format? Is it clean enough? Data is the most frequently underestimated prerequisite—and the one that causes the most project delays.
Technical complexity
Is an off-the-shelf solution available? Does it require customization? Fine-tuning? Development from scratch? Ready-made solutions have high feasibility; custom development has low feasibility.
Integration complexity
Are APIs available? Does the legacy system expose data? How many integrations are needed? Integrations with legacy systems that have no API are the biggest source of schedule delays in AI projects.
Organizational change
How many people need to change their process? Is there resistance? Who needs training? Organizational change is often the largest hidden cost of automation projects.
🟢 The four quadrants of the matrix
The combination of value and feasibility creates four clear decisions. The matrix makes explicit what was previously implicit—and forces the conversation every client needs to have about what to do first.
Quick Wins
High value + High feasibility
Top priority. Starting here delivers visible results quickly, builds trust, and creates political momentum for larger projects. Goal: deliver in 2-4 weeks.
Bets
High value + Low feasibility
Put it on the future roadmap. Invest when the prerequisites are in place — data is clean, the API is available, and the team is trained. Don't ignore it, but don't force it now.
Fill in
Low value + High feasibility
Do this only if you have capacity left after the quick wins. It’s easy to implement, but it doesn’t significantly move the business forward. Don’t let what’s “easy” dominate the roadmap.
Discard
Low value + Low feasibility
Don’t do it. Explicitly document why you ruled it out — “we ruled out X because its estimated value is low and integration would take 6 months” — so you don’t revisit the discussion.
⚡ Quick wins first — confidence and traction
Quick wins aren’t just “easy projects”—they’re a strategy for protecting large projects. Without a quick, visible win, long projects lose executive sponsorship before they deliver.
Characteristics of an ideal quick win
- ✓Deliverable and measurable in 2 to 4 weeks
- ✓A result visible to whoever approves the budget, not just whoever uses it
- ✓Data available and accessible without extensive preparation
- ✓Minimal organizational change — ideally, the new process is simpler than the old one
The credibility effect
A well-executed quick win changes the conversation: the client goes from “let’s see if it works” to “what else can we do?” It’s the difference between a one-off project and a long-term relationship.
⚠️ Bias toward overestimating value and underestimating effort
Planning optimism is universal and well documented. In AI projects, the bias is amplified by enthusiasm for the technology: the value seems enormous and the effort seems small — both are usually wrong.
Signs of value being overestimated
- ⚠"This will completely change how we do X" (before any data)
- ⚠ROI calculated assuming 100% adoption from day 1
- ⚠Included benefits that depend on other projects not yet approved
Signs of effort being underestimated
- ⚠"The data is ready" (without checking quality)
- ⚠Legacy integration estimated at “1 week” by someone who hasn’t accessed the system
- ⚠Team training not included in scope
💡 The 1.5× rule
Multiply every effort estimate by 1.5×. Reduce every projected benefit by 0.7× for the first year. Applied systematically, these adjustments produce estimates that come close to reality — and protect the consultant's credibility.
📊 Presenting the Prioritization to the Client
The prioritization meeting is where the consultant demonstrates the most value: not just by listing opportunities, but by recommend what to do first and why. That leadership is what sets consulting apart from research.
Prioritization meeting structure
Present the identified opportunities. 10 minutes.
Show the matrix with the opportunities positioned. Explain the criteria for each placement.
"I recommend starting with X and Y—here's why." Be direct; the client wants to be led, not consulted.
"Are there contexts I'm not aware of that would change this interpretation?" Incorporate them before finalizing.
Follow-up email with what was decided, what was ruled out, and the agreed next steps.
From analysis to decision
The completed matrix isn’t the deliverable—the decision made is the deliverable. Consultants who provide analysis without a recommendation waste their moment of greatest influence. Recommend. Justify. Decide together.
🎒 Module summary
Next track:
Track 3 — Audit: readiness, data, and risk