Why you should plan for role change in every role

Building career paths at an early stage company can be challenging. We unpacked this in our last post in this series and will continue to expand on it here, because itis easy to read workforce research and conclude that it applies to other companies, or to other people in the building. The McKinsey numbers say otherwise. More than 70% of workers could require some level of reinvention, defined as new tasks taking more than 15% of their time. About 47% could see 15 to 30% of their tasks reallocated, and another 25% could see 30% or more.

‍

The report is clear that this happens whether or not a job disappears: "most occupations reinvent regardless of whether employment grows, declines, or remains stable." Insurance underwriters are a useful example. Employment in the role stays flat at about 87,000 full-time equivalents, yet underwriters spend less time gathering data and more time assessing risk and managing AI. Same job title, different job.

‍

For an SMB, the practical consequence is that headcount planning and skills planning are separate exercises. You can have the same number of people next year and a different mix of work. Job descriptions written three years ago are probably already out of date.

‍

A task audit is the simplest way to find out where you stand. Take each role, or each team if roles are loosely defined, and do four things.

1.    List the ten tasks that take the most time.

2.    Mark which tasks a current AI tool could do, speed up,or check.

3.    Note which new tasks are appearing, such as reviewing AI output, handling exceptions, or answering customers who now arrive better informed.

4.    Estimate what share of the week those new tasks will take in twelve months.

‍

Any role where the new share lands above 15% goes on a watchlist. Those are the roles where training should start first, and the employees in them are the ones who will notice change first.

‍

The cashier example in the research shows what the new work looks like. Employment in that occupation drops from 2.9 million to 2.4 million full-time equivalents, and the work shifts from checkout toward customer interaction and AI oversight, with staff resolving "payment exceptions" that machines still handle poorly. The report also cites teachers using AI tools who saved an average of 5.9 hours a week, which they put toward direct student support.

‍

Time saved has to be directed somewhere. If you do not decide what the freed hours go toward, they will fill with more of the same work. Name the destination for each role: more customer contact, better quality checks, more selling, faster turnaround.

‍

Finally, the report cautions that "managing this load is increasingly a question for employers, not simply an individual responsibility." Change at 15 to 30% of a role is a lot to absorb on top of a full workload. Protect time for learning in the schedule, not in people's evenings.