Identify opportunities

Decide where AI adds value and where another approach works best

AI is powerful, but it may not suit every situation. This guide helps you decide where AI can provide real value or if another solution is better.

This stage builds on the business process map you created in Stage 1: Map your processes

Understand what AI can do well

Start by understanding what tools are available and what AI can help with. Then match the right tool to the right task in your mapped process.

Why this matters

If you choose the right tool for the right task, you’ll get better results with less rework. Knowing what AI can and can’t do helps you set clear boundaries and avoid surprises later.

Assess pain points for AI fit

Score each pain point from Step 1 using the criteria in Table 1.

Table 1. Criteria for scoring pain points for AI fit

Criterion  Low score (1)  Medium (2–3)  High score (4–5) 
Volume  Happens rarely  Weekly occurrence  Many times a day 
Repetitiveness  Varies each time  Some patterns  Consistent process 
Data availability  No digital records  Partial data  Data-rich
Error tolerance  Errors are catastrophic  Moderate consequences  Errors are easy to fix
Current cost  Low effort or cost Moderate resources  High effort or cost 

Why this matters

A simple score helps you compare options and quickly rule out ideas that are too risky, too costly, or not ready for your data.

How to do it 

Fill out Table 4 in the business process mapping template.

Example

After mapping the process with her team, Sarah scored each pain point to identify which ones were a good fit for AI.

Pain point Volume
(1–5)
Repetitive
(1–5)
Data 
(1–5)
Error Tol
(1–5)
Cost 
(1–5)
Total score
(AI fit)
Custom quote writing  20 
Document chasing  5 4 23
Document tracking  5 3 3 4 3 18
Xero data entry  4 5 5 3 3 20
Historical data import  3 3 4 2 5 17
Welcome email errors  4 5 5 4 2 20

Find opportunities and assess impact

Use the matrix in Figure 1 to prioritise each opportunity based on 2 factors:

  • how well it suits AI (your score)
  • what the business impact is.

Business impact means what happens if the outcome is wrong. Treat an opportunity as high impact if it:

  • could harm people or lead to unfair outcomes (for example, hiring or firing)
  • would be hard to undo once it happens.

If an opportunity has high impact on people, involve the people affected when you redesign the workflow (Step 3).

A 4-quadrant matrix showing how artificial intelligence opportunities are prioritised by business impact and AI fit. The left y-axis shows business impact (low to high), and the y-axis shows AI fit (low to high). The quadrants from top-left going clockwise are: Process improvement, AI opportunities, Quick wins, and Deprioritise.
Figure 1. AI opportunity assessment matrix

 

Low AI fit (score 5–12)

  • High impact – fix the process first, then reconsider AI.
  • Low impact – deprioritise for now (it likely isn’t worth the time or cost).

High AI fit (score 13–25)

  • High impact – prioritise for AI rollout.
  • Low impact – treat as a quick win for AI use.

These score ranges are a guide only.

Improve the process first

Not everything needs AI. If an opportunity has low fit for AI but high business impact, try simpler options first. For example:

  • templates – standardised forms, checklists or scripts
  • automation – simple rules-based workflows (such as email filters and scheduled tasks)
  • streamline – remove unnecessary steps
  • training – build staff knowledge and skills
  • integration – connect systems that don't currently share information.

Key rule: If a simpler solution can deliver 80% of the benefit at 20% of the cost or complexity, start there.

Why this matters

Simple fixes are often faster, cheaper and safer than introducing a new tool. They can remove the pain point without adding new risks. They can also help your return on investment.

Define scope and impact

Set clear boundaries for each opportunity you plan to improve with AI. For example, define: 

  • the intended use (and what is out of scope)
  • who is affected
  • what could go wrong
  • what would make it worth the time, effort and cost
  • what success looks like.

Why this matters

Clear boundaries prevent scope creep and reduce surprises later. They also make it easier to manage risk and explain how the tool will be used.

How to do it

Fill out Table 5 and Figure 1 in the business process mapping template.

Example

Sarah used the matrix to group each pain point by priority. For items with low AI fit, she listed simpler ways to fix the issue.

Priority level  Pain points 
Priority AI opportunities (high AI fit + high impact)  Document chasing (23), custom quote writing (20), welcome email (20) 
Quick wins (high AI fit + low impact)  Xero data entry (20) 
Process improvement first (low AI fit)  Document tracking (18) – try better checklist
Historical data import (17) – train a second bookkeeper

 

A 4-quadrant matrix highlighting priority opportunities such as document chasing and quote writing; and process improvements such as trying a better checklist first; to quick wins such as Xero data entry, and deprioritised items.

After scoring, Sarah and the team decided to: 

  • use AI to draft quotes and client emails 
  • use non‑AI automation for document tracking 
  • train staff to import historical data 
  • defer automating Xero data entry until the core process is stable. 

Sarah also wrote a short scope statement for each AI use case so the pilot can stay contained.

Download

Use this template across all stages to prepare your business.

Summary

You should now know which pain points are good candidates for AI and which are better solved another way. Come back and complete this stage for as many processes as you need.