Technology companies have spent the last two years telling themselves a comforting story about retention: that the market’s tight, that good people are hard to find everywhere, and that a bit of turnover is just the cost of doing business in a competitive sector.
That story is getting harder to hold onto. Recent industry research on AI-driven role transformation found something sharper than the usual retention headlines: attrition pressure is now building in sectors – financial services among them – that used to be comparatively stable, precisely because AI is redrawing roles faster than people can see where they’ll land in the new shape of things.
Employees aren’t just watching their pay and their manager. They’re watching whether their skills are still going to matter in eighteen months, and increasingly deciding not to wait and find out.
McKinsey’s research is blunt on what actually predicts attrition: a toxic or unclear culture is over ten times more predictive of someone leaving than compensation. That finding lands hard against how most organisations actually manage retention – with an annual survey and an exit interview, both of which tell you what already happened, not what’s building.
We heard the same gap described from the inside, in different language, at the Data Impact Leadership Forum in Edinburgh. The ownership question plainly: too often nobody actually owns retention – it sits in the grey area between HR, line managers and leadership, and that’s exactly where momentum quietly disappears.
Participants pushed further, into a blind spot most retention conversations skip entirely: confidence. The people organisations are most at risk of losing early, the group heard, are often the ones with real capability but not yet the self-belief or visibility to demonstrate it, invisible to a system that only measures what’s already been said out loud.
Recruitment doesn’t fix this either. Time-to-fill an external hire now runs roughly two and a half times longer than filling the same role internally, and the same research shows lack of growth and learning opportunity as the single most consistently cited driver of attrition, ahead of pay, ahead of almost everything else. Organisations are spending heavily to bring people in, while the people already inside quietly conclude nobody can see them growing.
The instinct, when retention becomes a board-level worry, is to build a better dashboard for leadership. That’s necessary, but it’s the wrong starting point. The gap FinTech Tables members described – capable people without visible evidence of their own progress – can’t be closed by a tool only managers look at. It needs to start with the individual: a place where someone can actually see their own development against their own goals, backed by evidence rather than relying on what a manager happens to remember at review time.
Only once that exists does the aggregate picture become honest. Delivery against goals, comparable across teams. Cultural alignment sitting alongside the output numbers, not as an afterthought. And the part most dashboards skip, a defined route for acting on what the data shows, so insight actually changes what happens next rather than sitting in a slide nobody revisits until the next annual cycle.
The Financial Services Skills Commission’s own 2026 research found firms increasingly “hoarding” scarce specialist talent – hiring into hard-to-fill technical and strategic roles while becoming less likely to backfill the people who leave. That’s a sector quietly admitting it can’t afford to lose the people it already has, at the exact moment AI is making those people more anxious about their own relevance, not less.
Annual engagement data was never built for a problem that moves as fast as AI. By the time a survey result flags a disengaged team, the strongest people on it have often already started looking. Retention risk builds continuously, in how people actually work and grow, which means it has to be visible continuously too – not once a year, and not only to the people at the top.
Our free 15-minute Leading through Disruption diagnostic, developed in partnership with SolasOS includes a dedicated section on talent visibility and retention – an instant, evidence-based read on where the gaps sit, for individuals and for the organisation as a whole.
Start the diagnostic: solas-os.ai/readiness
Next in this series: why the tools your technology teams bought aren’t the reason your AI initiatives are stalling.
This article draws on discussion from the Data Impact Leadership Forum, (Edinburgh, September 2025), alongside research from McKinsey, Fuel50’s State of Skills-Based Work (2026), and the Financial Services Skills Commission.
