AI automation can remove routine work while leaving employees unsure what their jobs are for. Procurify CFO Amy Wang argues that finance teams should map the work and discuss what replaces it before choosing a tool.
Watch Desk analysis
What happened
In a Forbes Council essay published on 27 October, Wang distinguishes between “transport work”, which moves information between people or systems, and “judgment work”, which involves decisions. She says agents are better suited to the former, but automating it can make people’s contributions less visible without answering what comes next.
Her practical suggestion is to trace a purchase from request to completion, ideally a non-catalogue purchase that crosses several teams. At each pause, record who handles it, how long it waits and whether the person is moving information or making a decision. Wang says her own test found roughly a dozen stops but only two genuine decisions, with copying, chasing and checking accounting for much of the wait.
Before buying a tool, she recommends discussing which transport tasks will go, what judgment work may grow, and what remains uncertain. If there is no concrete replacement work, she argues, leaders should say so rather than promise vaguely that staff will be freed for “higher-value” tasks.
Why it matters
A workflow map can show where automation might reduce delays, but it also makes the likely effect on roles harder to wave away. Wang’s advice is an executive’s account, not a guarantee that every finance team will find the same pattern. Still, it gives managers a more useful starting point than buying software first and treating confused adoption as an attitude problem.
Our read
The strongest part of Wang’s argument is also the least glamorous: follow one real piece of work, not the tidy diagram in the policy manual. If automation changes a job, people deserve a specific account of what changes and what does not. “Higher-value work” needs examples before it can earn the quotation marks.
What to watch
- Whether teams map real hand-offs before selecting automation tools.
- Which tasks are genuinely decisions, rather than information moving between people.
- Whether managers can name the work that will replace tasks they automate.
Discussion spark: Should employers have to map how automation will change people’s roles before buying a tool, or is that an unrealistic burden on teams still testing what the technology can do?
Sources and evidence
- Your Team Isn’t Resisting AI-They’re Waiting For You To Describe How It Impacts Their Jobs – Forbes (27 October 2026, 12:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.