Template

AI Opportunity Scanning Template

An interactive, scored template to capture, evaluate, and prioritize AI-driven opportunity candidates across products, services, and operations. Includes structured fields, a practical scoring rubric, a data-readiness checklist, and clear next-step recommendations.

Interactive Tool

AI Opportunity Scan

Purpose: Capture an AI opportunity candidate, assess value, feasibility, and risk, and produce a clear next step (pilot, research prototype, or deprioritize). This interactive template helps teams avoid hype-driven pilots by collecting consistent evidence early.

How to use: Fill each field with the best available evidence. Use the scoring fields to rate Value, Feasibility, and Risk. A suggested combined score and thresholds are provided below to help triage opportunities, but local context should guide the final decision.

Scoring guidance (suggested)

Value (1–5): Potential impact on revenue, cost, quality, throughput, customer satisfaction, or strategic differentiation.
Feasibility (1–5): Data availability and quality, modeling complexity, required infra, and organizational readiness.
Risk (1–5): Privacy/regulatory, safety, bias, reputational, or user harm—higher is worse.

Suggested combined Opportunity Score = (Value * 0.6) + (Feasibility * 0.35) - (Risk * 0.25). Scores roughly map: >3.6 = Good candidate for pilot; 2.6–3.6 = Research prototype / additional data work; <2.6 = Deprioritize or revisit later.

Note: This template records inputs for later comparison across opportunities and to feed experiment and governance workflows.

Unique ID (team use)

Optional short identifier (e.g., OPS-42). Helpful for linking to experiment trackers.

Opportunity title

Short descriptive name (one line).

One-paragraph summary

Describe the idea, who it helps, and the expected change. Keep to 3–6 sentences.

User or process pain

What problem does this solve? Who experiences it? How often and how severe is the pain? Include concrete examples or metrics if available.

Primary user / process segment

E.g., claims adjusters, warehouse pickers, online shoppers, machine operators.

Expected value (qualitative)

Describe expected benefits: cost savings, time saved, revenue uplift, quality improvements, risk reduction, or strategic value.

Expected value (quantitative estimate)

Numeric estimate in your preferred currency or units (annualized where possible). Enter 0 if not estimated.

Critical success metrics

List 2–4 measurable KPIs you would monitor (e.g., % defect reduction, minutes saved per case, incremental revenue).

Are usable data currently available?

Answer yes only if you can access representative data without major new integrations.

Data sources (list)

Where does the data live? (databases, files, APIs). Note owners and access needs.

Estimated usable sample size

Approximate number of records/transactions/events available for modeling or testing.

Data readiness checklist

Check items that are true for the available data.

Sufficient completeness (few missing values)

Timely (timestamps present and recent)

Reasonably accurate (validated against a source)

Required labels/ground truth exist or can be created

Class balance adequate or manageable

Access rights/permissions are or can be granted

Labeled data availability (scale)

1 = no labels; 5 = large labeled dataset ready for modeling.

1.0 10.0

Modeling complexity (scale)

1 = simple rule-based or logistic regression; 3 = standard ML pipeline; 5 = novel research or multimodal models requiring significant R&D.

1.0 10.0

Privacy / regulatory risk (scale)

1 = minimal/no personal data or regulated concerns; 5 = high regulatory scrutiny or cross-border data restrictions.

1.0 10.0

Privacy / compliance notes

List any regulations, required approvals, or specific concerns (e.g., HIPAA, GDPR, export controls).

Are there safety or bias concerns?

Consider physical safety, financial harm, reputation, or harmful model bias.

Safety / bias details

Describe potential harms, affected groups, and mitigation ideas.

Suggested prototyping approach

Choose the lowest-risk approach that still validates the core hypothesis.

Estimated effort (FTE-weeks)

Rough effort for a first prototype including data work, modeling, and evaluation.

Estimated incremental infra or tooling costs

One-time or monthly incremental costs (currency or units).

Key stakeholders and owners

List roles (e.g., product manager, data engineer, compliance, ops) and suggested owner.

Value score (1–5)

Rate expected impact: 1 low, 5 transformational.

1.0 10.0

Feasibility score (1–5)

Rate data and technical feasibility, plus organizational readiness: 1 low, 5 high.

1.0 10.0

Risk score (1–5)

Rate overall risk where 1 = low risk, 5 = high risk.

1.0 10.0

Combined opportunity score (manual calculation)

Use the suggested formula in the intro or your own. Suggested formula: (Value*0.6)+(Feasibility*0.35)-(Risk*0.25). Enter the result (range roughly -0.5 to 5).

Recommended next step

Select the recommended immediate action based on scoring and constraints.

Practical next steps and acceptance criteria for a pilot

List the minimum scope, success criteria, timeline, and owners for the chosen next step.

Additional notes / links

Links to data samples, dashboards, docs, or prior experiments.

You can explore this tool now. Sign in or create an account to save your responses and return to them later.

Make this tool part of your work

Save a personal copy, bring it to your team, or tailor the questions and workflow to fit what you are hungry to improve.

Member customization and team collaboration are coming soon.

Discuss Ask the Librarian

Discussion

Comments and conversation will live here.

AI Opportunity Scanning Template