AI Kick Start — Free Resource
12 questions to answer before your first AI project. Work through them with your team — honest answers now save wasted money later.
How to use this: tick each box once you can answer the question in one or two sentences. Eight or more ticks means you are ready to start a small pilot.
The best first AI project targets a measurable bottleneck, not whatever tool is in the news. If you can name the task and the hours it costs, you have a starting point.
AI can only automate work you can explain. If the process lives in one person's head, write it down first — that document becomes the brief for the build.
Pick one measure before you start: hours saved per week, faster response times, fewer errors. Without it, you will never know whether the project actually paid off.
Spreadsheets, inboxes, a CRM, paper forms — most businesses have all four. Scattered data does not stop a project, but it does add a tidy-up step you should plan for.
If yes, Australian privacy obligations apply. Know what you can and cannot paste into a tool before anyone signs up for a free trial, not after.
Decide the sign-off path now. Projects stall when approval is an afterthought — or worse, staff quietly adopt tools nobody has vetted.
A pilot without an owner quietly dies within months. Name one person responsible for keeping it working, fixing small issues and reporting on results.
Even an hour of hands-on training changes adoption dramatically. Without it, people drift back to the old process and the tool sits unused.
Have that conversation early and honestly. Projects framed as "removing the tedious parts of your week" get cooperation; projects that feel like surveillance get sabotaged.
Hours per week multiplied by hourly rate is close enough. This baseline is the only fair way to judge whether the project earns its keep.
Most useful AI tools cost between $20 and $200 per user per month. Setting a ceiling up front keeps the pilot honest and stops subscription creep.
Decide your exit criteria before you start: a date, a number, a review meeting. Stopping a failed pilot quickly is a good outcome, not a failure.
10–12 ticks: You are ready. Pick the smallest version of the project and start this month.
7–9 ticks: Close. Answer the gaps — usually data and ownership — before committing money.
6 or fewer: Hold off on tools. Map your workflow first; it is the cheapest fix on this list.