Algorithms now decide who gets the corner office
The next promotion cycle at your company might not hinge on a charismatic presentation or a golf-round endorsement. It will hinge on a score spit out by code that watched how long you hovered over a budget-allocation slider before clicking.
Boards from Siemens to Unilever have quietly handed slices of their succession-planning duties to third-party platforms that ingest everything—Slack metadata, CRM logs, 360-feedback sentiment, even mouse-heatmaps from training games. The output: a ranked list of ‘leadership material’ that bypasses the mahogany-lined discussion clubs where careers used to be made or buried.
The spreadsheet eats the old boys’ network
Traditional talent sessions were cocktail of memory bias and whoever told the best war story. Algorithms, hungry only for data, level that field—at least on paper. They flag that the high-potential finance director never delegates approval rights, or that the shy ops manager in Wrocław repeatedly delivers 18 % faster throughput during crises.
Yet the same code can’t smell fear in a workshop or register the way a plant floor quiets when a respected foreman walks in. It measures proxies: keystroke cadence, meeting bounce-rate, emoji choice. Feed it ten years of skewed appraisal history and it will happily mint a new generation of mini-me bosses who fit yesterday’s mold.
Black-box promotion, white-knuckle risk
European works councils are already pushing back. Deutsche Telekom’s labor board forced management to reveal that the vendor model downgraded candidates who took parental leave—because the pattern hurt a metric labeled ‘continuous project exposure’. The algorithm wasn’t sexist by design; it simply learned from an archive that equated presence with performance.
Regulators smell blood. Brussels is drafting rules that treat high-stakes HR analytics like medical devices: no CE mark, no deployment. Fines could reach 4 % of global payroll, not just revenue, if the model’s decisions can’t be explained in plain language to the worker it buries at rank 47.
Hybrids are winning—so far
Early adopters that keep humans in the loop report 30 % lower attrition among promoted execs, according to MIT Sloan data shared last month. The recipe: let the model surface hidden gems, then let people stress-test them for the unwritten stuff—moral compass, political acumen, ability to absorb public insult without flinching.
Ignore that two-step and you get what one Fortune 100 CTO calls ‘brilliant jerks with great GitHub stats who can’t chair a five-minute stand-up without mutiny’. The board fired three data-blessed rising stars in twelve months; shareholder questions followed.
The takeaway isn’t man versus machine—it’s man plus machine versus panic hiring. The firms racing ahead aren’t those buying the shiniest SaaS dashboard; they’re the ones rewriting promotion criteria so that empathy and ethics carry numeric weight before the code ever crunches a byte. Because when the next downturn hits, the algorithm won’t defend its pick. The board will—and by then the spreadsheet will be long since deleted.