Ask a finance director what they want from their next FP&A hire and the answer has shifted noticeably in the past two years.
Modelling ability is assumed. What they are describing instead is someone who can explain what the numbers mean to people who don’t work in finance, and who can be trusted to tell the business something it would rather not hear.
That change is being driven by automation, though not in the way the headlines suggest.
Where finance actually is with AI
Adoption is further behind than the noise implies. The FP&A Trends Survey found that 53% of organisations still use no AI at all in any FP&A process. Research summarised in the State of AI in Finance 2026 report put a figure on why: 68% of CFOs said they had been slow to adopt because they did not know where to start.
Capability, meanwhile, has become the constraint finance leaders worry about most. A Gartner survey of 100 CFOs conducted in early 2026 found that building AI and digital talent in the function was one of their two most challenging near-term priorities.
So most finance teams are in the same position: the tooling is available, the ambition exists, and the people who can bridge the two are scarce. That combination is what is reshaping the FP&A brief, rather than any wholesale replacement of finance roles.
What that means for the roles you are hiring now
Less time on assembly, more on interpretation
The parts of FP&A that automate first are the parts that consume most of a junior analyst’s week: pulling data, refreshing reports, reconciling variances. What remains is harder. Deciding which variance matters, understanding the commercial cause, and recommending an action.
That changes what a good hire looks like at every level. A candidate who is excellent at building a model but uncomfortable in front of a commercial audience is a narrower hire than they were three years ago.
Judgement over output volume
When a forecast can be produced in minutes, the value sits in knowing whether to believe it. Finance hires increasingly need to interrogate an automated output, spot where an assumption has drifted, and be confident enough to override it.
Ownership of data quality
FP&A leaders are inheriting responsibility for the integrity of the inputs, not just the elegance of the outputs. At manager or head-of-FP&A level, governance of models and data is now part of the role, whether or not it appears in the job description.
How to interview for it
Four questions that surface the difference quickly:
- “Talk me through a forecast you got wrong.” You are listening for how they diagnosed it, not whether they have one.
- “Explain a variance to me as though I run the sales team.” Clarity under mild pressure tells you more than a technical test.
- “Where would you not trust an automated output?” Candidates with real exposure have specific examples. Those without speak generally about accuracy.
- “What did you stop doing to make room for something more useful?” Prioritisation is the skill most finance teams under-assess.
A note on the internal pipeline
If automation absorbs entry-level work, the traditional route into commercial finance narrows. Businesses that plan for this, by giving junior analysts business-partnering exposure rather than only reporting work, will have a senior pipeline in three years. Those that do not will be buying it on the open market at a premium.
Talk to Insight Select about your finance hiring
We recruit across finance and commercial functions, and we spend a good deal of our time helping clients define what they actually need from an FP&A hire before the brief goes out.
