There is no single right span of control for engineering managers, but common guidance for knowledge work lands around 5 to 10 direct reports, and Gallup puts the median at about five to six (Gallup). Engagement tends to drop as spans grow past that, and past roughly 15 reports, managers struggle to run real 1:1s at all.
When a company grows fast, manager spans quietly inflate. A lead who had five reports has nine, then thirteen, and nobody decides this, it just happens between reorgs. The research on span of control suggests that drift has a real cost, and that it shows up first in the things nobody measures: 1:1 quality, career development, and how well anyone actually knows what's going on.
Key Takeaways
- Gallup reports the average manager's span rose from 10.9 to 12.1 in a year, against a median of about 5 to 6 (Gallup on span of control).
- Managers with 7 or fewer reports score meaningfully higher on team engagement than those managing 15 or more (span-of-control data).
- Newer managers do better at the low end (5-8); experienced managers stretch further.
- Past roughly 15, managers struggle to deliver quality 1:1s and development at all.
What the Research Suggests
Gallup's data captures the drift directly: the average number of people reporting to a manager climbed from 10.9 to 12.1 in a single year, while the median sits at roughly five to six (Gallup). That gap between a low median and a rising average is the story: most managers have a handful of reports, and a growing tail are carrying far more. Gallup is careful not to name one optimal number, arguing that larger teams succeed only when organizations invest in the conditions that let managers lead them. Practitioner guidance for knowledge work typically lands around five to ten, because coaching, unblocking, and career conversations don't compress well.
The engagement effect is the striking part. Managers with seven or fewer direct reports score notably higher on team engagement, on the order of 20% higher, than those managing fifteen or more (the attention math). Beyond about fifteen, the bandwidth problem becomes obvious: 1:1s get cancelled, development conversations disappear, and the manager becomes a bottleneck rather than a multiplier. Reported spans at large tech companies tend to sit in the mid-to-high single digits, which matches the pattern.
| Span | What tends to happen |
|---|---|
| 5–8 | Close mentoring, strong visibility (good for newer managers) |
| 8–12 | Workable for experienced managers |
| 12–15 | Development starts getting squeezed |
| 15+ | 1:1s slip, coaching disappears, manager becomes a bottleneck |
Why Wide Spans Fail Quietly
The failure mode is subtle, because a manager with eighteen reports still looks busy and the org chart still looks fine. What degrades is invisible on a dashboard. Each person gets less attention, so problems surface later. Career conversations get deprioritized because they're never urgent. The manager defaults to firefighting whoever is loudest, which means the quiet, competent people get no attention at all until they resign.
This compounds with the finding that managers account for a large share of the variance in team engagement. If manager quality is that decisive, then stretching managers past the point where they can actually manage is an expensive way to save a headcount.
Remote roles for engineering managers. Ruzora hires experienced engineers and managers from LATAM for remote work with US teams. If you are the manager with nine reports and climbing, see the open Engineering Manager role, or the Founding Engineer role if you would rather be back in the code.
A Concrete Version
A startup grows from 12 to 30 engineers in a year without adding managers, so two leads end up with fifteen reports each. Nothing dramatic breaks. But 1:1s slide to every three weeks, then "as needed." Two strong engineers quietly disengage and leave within a quarter, and in the exit conversations both say some version of "I didn't know where I stood or where I was going." Neither departure gets attributed to span of control, because that's not a metric anyone was watching. Adding one more manager would have cost far less than replacing two senior engineers.
The Honest Counterpoint
Narrow spans have real costs too, and "fewer reports is always better" is wrong. Very small spans create management-heavy orgs with too many layers, which slows decisions, adds coordination overhead, and can leave managers under-occupied and inclined to micromanage. The right number also depends heavily on context: a team of experienced seniors doing well-understood work needs far less attention than a team of juniors on ambiguous problems, and a manager who also carries significant technical or strategic work has less capacity for people. Treat 5-10 as a default to reason from, not a rule to enforce.
What This Means for Teams
Span of control is worth tracking as deliberately as headcount, because it drifts silently and its damage shows up as attrition rather than as a metric. Two practical habits: notice when a span crosses roughly ten and treat that as a trigger to plan, rather than waiting for something to break; and remember that a manager doing hands-on technical work effectively has a smaller capacity than the number suggests, which is part of what the engineer/manager pendulum makes explicit. See available engineers.
A stretched manager is also the last to notice someone checking out; these are the early signs a developer is about to quit.
Frequently Asked Questions
How many direct reports should an engineering manager have?
Practitioner guidance for knowledge work lands around 5 to 10, and Gallup's median sits near 5 to 6. Newer managers do better at the low end (5-8), experienced managers can handle more, and past about 15 the role starts breaking down.
What happens when spans get too wide?
1:1s get cancelled, career development disappears, and the manager defaults to firefighting the loudest problems. Managers with 7 or fewer reports score meaningfully higher on team engagement than those with 15 or more.
Can spans be too narrow?
Yes. Very small spans create too many layers, slow decisions, add coordination overhead, and can encourage micromanagement. The goal is a sensible default rather than minimizing the number.
Does the type of work change the answer?
Considerably. Experienced seniors on well-understood work need less attention than juniors on ambiguous problems, and a manager carrying technical work has less real capacity than the headcount implies.
The Bottom Line
Manager spans inflate silently as companies grow, and the damage shows up as disengagement and attrition rather than on any dashboard. Gallup's average span is climbing while its median stays near five to six, and common guidance for knowledge work lands around 5 to 10 reports, with engagement dropping as spans widen past that. Treat crossing ten as a planning trigger, adjust for how much technical work the manager carries, and remember that adding a manager is usually cheaper than replacing the people a stretched one loses.
Managers and tech leads: see the open engineering manager, founding engineer and forward deployed engineer roles.
Roberto Espinoza is CEO of Ruzora, which helps US startups hire pre-vetted senior LATAM engineers, with a vetted shortlist in 72 hours. See available engineers.
