How It Works
A proven 4-step framework to ensure you make informed decisions based on a clear understanding of what the risks are and where the value actually lives.
Start with the problem, not the product.
I cannot stress enough how important it is to start with solving a problem you are having right now. The best way to learn AI is to use it for something that will positively impact your business in the next 90 days. You don’t want it to feel like something extra you have to do on top of everything else you need to do.
AI is ultimately a tool, not a strategy in and of itself.
Pick a current business problem
Growth, Capacity, Efficiency, Quality, Insight
Identify opportunities for improvement
Bottlenecks, Inefficiencies, Personalisation
Evaluate Feasibility
Skills, Data, Technology, Budget, Risks
Define what success looks like
Measure
Now it’s time to do the hard work.
The difference between successful and unsuccessful AI adoption isn’t about the technology itself, it’s about preparation. Too many businesses rush to implement AI tools without first addressing the foundations that make AI effective. AI requires clean data to produce accurate results, documented processes to automate reliably, trained staff to use it properly and clear guidelines to use it responsibly.
Document the process
Include edge cases. Identify the steps that AI should do.
Clean the data
Focus only on the data required for the pilot.
Choose the tool
Look at the tools you already have available to you. Use off-the-shelf tools wherever possible.
Train your staff
Give your staff the best chance of success by ensuring they have the necessary skills.
Communicate the plan
Ensure all staff know that AI is not being introduced to replace jobs.
Now the fun begins.
Expect to experiment A LOT. Try different ways of asking the same question, run the same task multiple times if the first attempts are wrong, test different approaches, make mistakes, backtrack, and try again. Be patient: research shows it takes about 11 weeks to see consistent productivity improvements, so budget for this learning curve.
Commit for 90 days
Don’t scale too early. Budget in the learning curve.
Experiment
Use this time to learn. Test the limits.
Review all AI Output
Make sure there is a “Human in the loop”.
Improve through iteration
Research shows it takes about 11 weeks to see consistent productivity improvements.
Document your findings
Strengths/Weaknesses
Don’t scale before you validate.
Scaling is rarely just a technology challenge. It is mostly a change management and governance challenge.
Evaluate pilot’s performance
Compare results to success criteria set at the start of the project.
Create an acceptable use policy
Clearly document what employees are allowed and explicitly not allowed to do with the AI tools.
Develop role-specific training
Ensure staff have the necessary skills to get the best results.
Implement a Feedback Mechanism
Give staff a way of logging issues and ideas to improve the workflow.
