Building an AI-ready planning team: skills, careers and the first 90 days
The first four parts of this series looked at the work, the capacity, the vendors and the rules. This one is about the team that has to cope with all of it: what to hire for, how to develop people, and what to do on Monday morning.
You will be asked
In Gartner’s late-2025 survey of 321 customer service and support leaders, 91% said they were under executive pressure to implement AI. Whatever sector you plan for, expect the question “where is the AI in planning?” sooner or later. A team with a considered answer, including the parts it has deliberately not automated, is in a much stronger position than one improvising a reply in a budget meeting.
The skills that keep their value
The craft underneath planning doesn’t expire: queueing intuition, an understanding of variance and bias, contact-reason thinking, and the ability to reason about a schedule intuitively. These are the skills that let someone look at an automated output and know it’s wrong. They are also the skills most at risk of atrophy if the tool does all the work, so deliberately protect them: have juniors build a forecast by hand once a quarter, and have the team explain an Erlang result without the calculator.
The skills to add
Data fluency: enough SQL, Python or spreadsheet modelling to interrogate the data behind a result, not to become a data scientist. Backtesting and experimentation: the habit of holding data back, comparing a new method with a baseline, and reporting bias as well as error. Verification discipline: checking AI-generated analysis the way you would check a new starter’s, with reasonableness tests and spot checks. Vendor evaluation, as described in part three. Governance and change: the decision register from part four, and the ability to explain it to staff and unions. And, as ever, commercial storytelling, because the value of the planning function still has to be argued to finance. Skills that move you up in workforce planning and communication versus technical skills go deeper on the people side.
The junior pipeline problem
The tasks that automate fastest are the ones juniors have always learned on: pulling data, building the first-cut forecast, assembling reports. Trade commentary has noted that as AI reduces entry-level work, organisations face pressure to redesign early-career development. The planning function has the same problem. If the apprenticeship disappears, so does the supply of experienced planners in five years.
One answer is to change what juniors do rather than whether they exist. Make the verification role theirs: reviewing automated forecasts and schedules, investigating the exceptions, and writing up why the model was wrong on a given day. It’s a faster route to judgement than building the baseline ever was, as long as someone senior reviews their reasoning. For career routes in and out of the function, see the workforce planning career ladder and moving from agent to planning analyst.
The first 90 days
Days 1 to 30: inventory and measure. List your team’s tasks and apply the three-question test from part one. Record your current accuracy (WAPE and bias) as the baseline. Find out what AI tools people are already using, approved or not. Start measuring where AI changes the work, not just whether it saves time: Deloitte’s UK data suggests workers report saving about 70 minutes a week, only half of users say it saves them time, and most who do save time reinvest it in more work, so time saved isn’t the same as capacity released.
Days 31 to 60: one pilot, properly designed. Pick one task that passes the test (a baseline forecast or schedule suggestions is a good candidate). Write down what success and failure look like before you start, hold back a comparison period, and keep the human benchmark running. The eight-step implementation framework gives you the gates.
Days 61 to 90: set the rhythm. Stand up the decision register, agree who owns AI-related planning risks, and give your operations director a one-page scorecard: forecast accuracy against the benchmark, hours reallocated, exceptions caught, escalation behaviour in the hybrid queue, and a view from the people affected. How a planning team can show its success covers the wider reporting habit.
The closing thought
The evolving role of AI in workforce planning isn’t a story about replacement. It’s a story about where the judgement sits. Teams that know which decisions they have handed to software, can show the evidence, and have kept the skills to overrule it will do well. Teams that haven’t will find out the hard way on a busy Monday.
See also
- The ops leader’s AI stance — invest, refuse, retire
- The minimum viable planning team — by operation size
- Adjacent careers from workforce planning
- From forecaster to orchestrator: how AI is changing the planner’s job
Sources and further reading
External sources used in researching this article, accessed October 2026. Vendor-authored and trade-press material is identified as such; figures are quoted as reported and have not been independently verified.