A reusable reference for identifying where AI may create meaningful value, defining practical use cases and turning those decisions into an implementation roadmap.
Core idea
Opportunity discovery is not a list of AI ideas. It connects business direction to real workflows, compares value, viability and velocity, and turns decisions and dependencies into an implementation roadmap.
Where this guide fits: Guide 01 establishes the implementation foundations. Guide 02 turns business direction into prioritized use cases and a roadmap for the knowledge, governance, connections, collaboration, operating, and adoption work that follows.
Start here
Prepare what you already know
Review the guide and identify a few areas of work worth discussing. Bring whatever is useful: examples, process documents, notes, a completed worksheet, emails, data or evidence from earlier AI experiments. Perfect preparation is not required.
Work through it together
Examine the opportunities with GrowthIQ
We will confirm priorities, walk through real workflows, clarify the intended role of AI and identify missing decisions, dependencies, owners and evidence. Some questions may be resolved during the conversation; others will become roadmap work.
Build and maintain the roadmap
Turn the evidence into an operating plan
GrowthIQ and NAVI will organize the available evidence into defined use cases, open questions, dependencies and next actions. We will review the roadmap with you and continue updating it as decisions are made and implementation evidence develops.
This is an ongoing process rather than a one-time workshop. Early opportunities can move forward while the organization prepares knowledge, governance, connections and other foundations needed for later work.
01DirectionWhich business outcomes and areas of work matter most?
02Use caseWhich workflows could materially improve?
03PriorityWhat deserves action now, preparation next or no further investment?
04RoadmapWho owns the work, what must be resolved and what evidence guides the next decision?
Start with what the organization is trying to change
Business direction establishes context before the organization discusses tools, agents or features.
What is the organization trying to improve, protect or create?
Which business areas are under the greatest pressure or have the greatest potential?
Is the priority speed, capacity, quality, growth, customer experience, risk reduction, innovation or something else?
Who owns the direction and can resolve cross-functional decisions?
What constraints or commitments already shape the roadmap?
Opportunity lensCreate new capabilityEnable a new service, decision, experience, product or way of operating.
The most useful participants usually include a business sponsor or decision maker, the person accountable for the workflow, and one or more people who perform the work. Add technical, security, legal or governance participants when the opportunity requires their decisions.
Define use cases from real work
A useful use case begins with recurring work that has a recognizable trigger, participants, decisions and completed output. Observation, examples and existing outputs are often more useful than brainstorming alone.
Outcome and ownerWhat result should the workflow produce, and who is accountable for it?
Current workWhat triggers it, who participates, what decisions occur and what output completes it?
Friction and baselineWhere do delay, searching, rework, errors, inconsistency or missed opportunities occur?
AI contributionShould AI retrieve, create, analyze, recommend, coordinate or take a defined action?
People and work impactWhich tasks may be reduced, accelerated or changed; which responsibilities remain; what new work appears; and what learning or support will people need?
Inputs and systemsWhich knowledge, data, communications, tools and connections are required?
ConditionsWhat permissions, human review, adoption, process changes, dependencies or exceptions must be addressed?
Not sure where to begin? Look for repeated searching, manual preparation, recurring questions, bottlenecks, heavy handoffs, inconsistent outputs, avoidable rework or decisions that require information from several places.
Create a concise use case brief
Workflow and business area
Owner and intended users
Current problem or opportunity
Proposed role for AI
Intended outcome and baseline
Required knowledge, data and systems
Human decisions, permissions and boundaries
Tasks changed, retained responsibilities and support needs
Intended use of capacity and new work created
Dependencies, risks and potential reuse
Use value, viability and velocity
Prioritization compares each candidate use case through the same three lenses, then adjusts the emphasis to reflect the organization’s priorities, constraints and risk.
01
Value
If this works, how much will it matter to us?
Consider strategic importance; time, cost, quality, capacity, growth, customer or risk impact; frequency, reach and reuse.
02
Viability
Can we make this work responsibly?
Consider workflow clarity, ownership, knowledge and data, permissions, risk controls, technical feasibility and change readiness.
03
Velocity
How quickly can we test it in real work and learn?
Consider time to a useful version, access to users and inputs, effort, speed to observable results and ability to revise.
Compare candidates relative to one another rather than against a universal passing score. A simple high, medium or low assessment is often enough for the first conversation. The emphasis changes with context: regulated work may prioritize viability, while a stalled initiative may need faster evidence.
Give every use case a clear portfolio status
Before implementation, keep ideas visible without turning every idea into an active project.
Status
When it fits
Pilot now
High value, sufficient viability and a realistic path to useful near-term evidence.
Prepare next
The use case is understood and potentially valuable, but specific ownership, knowledge, access, governance, adoption or integration gaps must be resolved.
Approved self-service
Useful, fast and low-risk work that a person or team can pursue within approved tools, policies and information boundaries.
Explore later
The potential may be meaningful, but more discovery is needed to understand the workflow, value or responsible implementation approach.
Decline or redirect
Low value, poor solution fit, unclear ownership, uncontrolled consequences or a problem better solved another way.
Outcome and rationaleWhy the item matters and why it has this priority.
Owners and usersWho is accountable, involved and authorized to decide.
Scope and boundariesIncluded work, exclusions and permitted AI actions.
Dependencies and changeKnowledge, data, systems, governance, adoption, workforce support and process requirements.
EvidenceBaseline plus value, quality, adoption and risk measures.
TimingResponsible parties, target dates and blocking decisions.
Next decisionScale, revise, prepare further or stop.
Lessons and reuseWhat should be preserved for future workflows.
A foundation gap does not automatically eliminate a use case. Missing knowledge, permissions, ownership, connections or governance usually becomes explicit roadmap work that affects scope, sequence and timing.
Choose a meaningful first pilot
The first pilot should use a real workflow, produce useful evidence and leave reusable capability without requiring the full future-state system to be built.
Narrow scope. Meaningful ambition.
Define the user group, included inputs and systems, permitted AI actions, required human decisions, expected output, exclusions, timeline and scale/revise/stop criteria before testing begins.
Meaningful recurring problem
Accountable owner and available users
Recognizable trigger and output
Trustworthy and permitted inputs
Manageable initial scope and risk
Measurable result or learning objective
Appropriate human decisions
Path to reuse or adjacent work
Example: client onboarding preparation
ValueFaster preparation, more consistent plans and less repeated senior support.
ViabilityOwner, examples and reviewers exist; source ownership and access need confirmation.
VelocityA small set of completed or low-risk engagements can be tested within weeks.
RoadmapPilot now; prepare source improvements and a system connection as later items.
Make the post-pilot decision
After testing in live work, review output quality, corrections, time, exceptions, adoption, support burden, failures and business results. When no reliable baseline exists, establishing one can be an early pilot objective; useful evidence may be quantitative or qualitative.
ScaleUseful, sufficiently reliable, owned and repeatable.
ReviseThe use case matters, but the scope, workflow or AI role should change.
PrepareValue is visible, but knowledge, controls, training or integrations need work.
StopValue is insufficient, adoption is weak, risk is unmanageable or another solution is better.
Prepare for the roadmap conversation
Bring what you know and use the gaps. Missing information is not a failure; it becomes part of the roadmap. The discussion is most productive when someone can speak for business priorities, workflow ownership and the day-to-day work.
Current business priorities and accountable leaders
Areas experiencing pressure or opportunity
Two to five workflows worth discussing
Examples of current inputs and outputs
Known process owners and intended users
Available baseline or performance evidence
Knowledge, data, access and system dependencies
Risk, approval or policy constraints
Existing AI experiments and lessons
Decisions requiring leadership input
Adapt the framework to the organization’s industry, size, obligations, risk tolerance and operating model. It is not legal, regulatory, security or industry-specific professional advice.