Loading today’s stories
Leadership1 distinct publisher3 min readPublished
Managers carrying 20 reports and five hours of daily meetings are still expected to ship code, so the development hours went first. The bill arrives later, in retention and a thin internal bench.
The Board Room · Leadership desk
Compiled by The Board RoomSomething wrong?How this is made
Follow any of these and your For You feed starts watching them — no settings page required.
leadership
Disney swaps raises for discounted stock and a full health-plan re-enrollment1 distinct publisher
leadership
Mamdani's tax office says the data was always public, and sent a memo instead of a witness1 distinct publisher
leadership
Robotaxis Are Taking Mid-Teens Share in Three Metros. Headcount Will Not Show It.1 distinct publisher
leadership
Sponsorship Is Not Retention: What An Amazon H-1B Win Still Does Not Buy1 distinct publisher
Start with the arithmetic a span imposes. Roughly 20 reports and about five hours a day in meetings works out to some 25 meeting hours a week, and a modest half-hour with each report would claim 10 of them, close to 40 percent of the time Mitchell Kosowski already spends in rooms [5][19]. That is before the code he is also expected to write [5]. Nothing in that budget forbids coaching, which is why he says he protects it [6]. It does mean coaching competes every week against work that has a ship date, and only one of the two shows up in a sprint review.
Deel's published target is the cleanest instrument we have for the mechanism. The company used to aim for eight to 10 employees per manager and now aims for 12 to 15, which moves the midpoint from nine to 13.5, about 50 percent more reports per manager [10][17]. Hold the manager's development hours constant and each person's share falls by roughly a third [18]. Cofounder and CEO Alex Bouaziz attributes the change to AI absorbing administrative work that once capped team size [10]. The administrative load is genuinely compressible; the hour you spend on someone's third year is not, and it is the one that scales linearly with headcount.
The record also contains its own ceiling. Bouaziz says handing a manager 20 to 30 direct reports amounts to saying you do not want interactions between the manager and individual contributors at all, and that the layer of managers of managers is meant to disappear [13][12]. Kosowski, at a different company, is already at about 20 [5]. Stated span policy and lived span are drifting apart inside the same industry, and the target number tells you less than the escalation path underneath it.
A skeptic has a good answer here, and it is in the source. Paddy Lambros of Dex points out that the old model promoted star performers into management to scale impact without asking whether they wanted to coach or could [8], and that AI now makes it viable for those people to stay as senior individual contributors, which is how Dex staffs former founders and heads of engineering [9]. That fixes a misassignment, though it does not retire the demand for coaching itself. Spellbook's Scott Stevenson describes what replaces the formal channel: on-the-fly, in-the-trenches feedback folded into the work [15]. Useful, and biased toward whoever is standing closest to the work. The one-to-one is what gets cut first when someone needs an extra 30 to 60 minutes, and it is the one place the non-urgent question of where a career is going gets asked [16]. Lambros's own summary of the new arrangement is sink or swim [4].
Bouaziz names the deferred cost without quite pricing it: the leaders of tomorrow are the people you cared for a long time ago [14]. That is the sequencing problem in one line. The decision made this quarter is to widen the span and ship faster; the consequence lands two and three years out as external senior hires, longer time-to-fill on lead roles, and the attrition of people who could not see a next step.
The record so far supports the mechanism, not the size of the bill. This is testimony from three executives and one engineering manager, with no attrition, promotion-rate, or tenure figures attached [20]. The mechanism is legible, but nobody has measured its magnitude. Any company widening spans this year is making a wager on it either way, and the honest version of the board slide says so rather than booking the manager reduction as clean savings.
Ranked by verification strength, evidence, and original report placement.
Business Insider reports that across tech, the "Great Flattening" has meant fewer managers overseeing larger teams and racing to keep up as the industry remakes itself around AI.
The result, per Business Insider, can be less time for managerial work that does not produce an immediately visible result, such as coaching employees on where their careers are going.
Paddy Lambros, founder and CEO of Dex, which uses AI to connect software engineers with employers, said the days of "purely pastoral managers" running teams of five or six are over.
Lambros said managers are increasingly coding themselves and the pace makes relationships more about getting work done: "It's much more tactical and transactional. It's not managing the human, it's managing the work. And it's much more sink or swim."
Mitchell Kosowski, a software engineering manager at a tech startup, said he has about 20 people reporting to him, spends about five hours a day in meetings, and is also expected to code.
Kosowski said "The joke is I code 90% of the time, and I manage 90% of the time, which I think doesn't feel like much of a joke these days," and that he tries to protect time for coaching.
Distinct publishers with included, body-backed reporting in this cluster.
businessinsider.com
1 article · August 28, 2026
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One outlet, four voices, no measurements
Everything traces to Business Insider's own interviews: three founder-CEOs describing companies they run, plus one engineering manager describing his week. Two figures are specific enough to argue with — Deel's move from eight-to-ten reports to twelve-to-fifteen, and Kosowski's roughly 20 reports against five hours of daily meetings — and neither has been checked by anyone else. Nothing in the reporting measures what wider spans do to the people inside them.
Three companies, three concrete changes
Deel has actually shifted its target span, Dex staffs senior people as individual contributors instead of managers, and Spellbook keeps annual and midyear reviews as the formal backstop for careers. That is practice rather than prediction — but it is three software vendors, disclosed by their own chief executives, with no survey, no headcount data and no indication of how common any of it is across tech.
The framing outruns the arithmetic
'The Great Flattening' enters as settled industry fact, yet the only span number belongs to one HR platform, and Bouaziz himself calls 20-to-30 reports the point where interaction stops — roughly where the engineering manager the story treats as typical already sits. Our own promise that the bill lands later, in retention and a thin internal bench, is inference: no attrition, promotion or tenure figure appears anywhere in this reporting, and the per-report time loss we compute is arithmetic on a target, not an observed hour.
Three of four voices sell the trend
Lambros runs a company that uses AI to place engineers, Bouaziz runs an HR and payroll platform, Stevenson sells an AI contract tool — each has product downstream of a world where AI reorganizes management, and each is describing his own org chart. Kosowski, the manager actually carrying 20 reports, is the only person here with nothing to sell, and he supplies the least flattering detail in the piece. Business Insider signs off by inviting readers to share their experience in the comments, which is engagement, not corroboration.
Firm on what was said, soft on what it means
We are confident in the quotes and the two numbers behind them; we are much less confident in the trend they are mounted on. A single outlet, no independent confirmation, and executives characterizing their own companies leaves the direction plausible and the magnitude genuinely unknown — including whether coaching time was ever budgeted in the first place.