The Subtraction Promotion

If you promote your best engineer into their first management role and leave their full workload intact, every hour spent managing competes with work you’ve already promised. When they write less code, are they failing the new job or finally doing it?

The move looks like addition: a team that needed a manager now has one. Count again. You’ve taken your strongest contributor out of full-time delivery, given them a job they’re still learning, and left the old work assigned to them. The org chart shows a promotion. The capacity plan counts the same person twice.

That’s the subtraction promotion. You hope their leadership will eventually improve the team’s output enough to repay the cost of moving them. But someone has to absorb their old work, and learning to take it over consumes time too. Counting the hoped-for gain immediately leaves the transition unfunded.

If the engineer already mentors colleagues and coordinates delivery, a promotion may recognize work already in the plan. A hands-on lead can be a sensible fit for a small team. Count the new responsibilities, though: responsibility for a colleague’s performance may require a difficult conversation that helping them debug never did. The test is whether the combined work fits, not whether the lead still writes code.

In their 2019 account of research at 131 firms, Alan Benson, Danielle Li, and Kelly Shue found that higher sales predicted promotion. Among those promoted, however, higher previous sales predicted a weaker contribution to their subordinates’ sales. A doubling of pre-promotion sales corresponded to a 7.5% decline in estimated manager value added: the contribution the manager made to subordinate sales, after accounting for worker ability and other factors. That estimate concerns sales management in this sample, rather than a forecast for your new engineering lead.

The authors’ explanation makes the finding more useful than a warning against promoting good sellers. Top sellers appeared to face a lower bar for managerial ability. A weaker seller had to demonstrate more managerial potential to get promoted at all. The researchers also found that collaboration experience, measured through shared credit on sales, predicted better management performance without making promotion more likely.

They leave room for a reasonable motive. A promotion policy that rewards sales can encourage people to sell, even if it produces weaker managers. The authors point to pay incentives and separate career paths for individual contributors as ways to reduce that conflict. Rewarding excellent work and choosing someone to manage it deserve separate decisions; development can’t substitute for evidence that the person wants and can learn the new job.

The study examines selection and managerial performance. Our additional concern is what happens after you choose: even a promising first-time manager needs room to practice. Leaving their old workload intact makes the unfamiliar work compete with the work that earned them the promotion.

Suppose the new lead owns an integration needed for the next release. They hand it to another engineer, who needs help understanding the design. While they work through it together, the plan needs to count both people’s time. If the deadline stays fixed and the lead remains responsible for rescuing delivery, doing it themselves may be the fastest way to meet it. You’ve given them a reason to keep using the skill you promoted them out of.

A handoff needs a named recipient and a decision about what that person stops doing. It also needs a boundary around the lead’s involvement. If every design choice still requires their approval, replacing their name on the ticket releases less time than the plan may assume. Agree which choices the recipient can make and where the lead must review; keep that review and teaching time in the plan.

Where nobody can absorb the work, something has to change in the delivery commitment or the staffing. Keeping the integration with the lead for a critical release may be the right trade. Then the management work it displaces needs an owner, too. Calling the role a promotion doesn’t settle any of those choices.

There is evidence for funding the learning, too. A 2017 meta-analysis of 335 independent samples by Christina Lacerenza and colleagues found positive average effects of leadership training on learning, application at work, and results. Its analyses favored feedback, practice, and training sessions spread apart. The research covers leadership training broadly; it doesn’t establish that a particular coaching arrangement will fix a poor appointment.

For a new lead, that points toward support around the decisions they’re actually making. Rehearse the hard feedback before they give it, then discuss what happened afterward. Work through the moment when doing the task themselves would be faster, but someone else needs a chance to learn. Agree which delivery risks they can accept while that person learns, so delegation has your backing when it costs something.

Start with the new lead’s own manager. If they have the experience and can make time for this work, use that relationship. An experienced internal peer can help with a gap in skill; an outside coach is an option when useful support is unavailable internally. In either case, the new lead’s manager still owns the workload and has to back the authority they’ve assigned. Buying coaching doesn’t reassign a single ticket.

Development also needs evidence of progress: a difficult conversation the lead can now handle, a decision the team can make without waiting, work completed without the lead taking it back. Pair that evidence with the workload they actually carried. If delivery emergencies consumed the time set aside for management, revisit the plan before treating missed management work as evidence of poor fit.

If the lead has that room and support but still avoids essential management work, address the performance gap. Training sessions attended can’t stand in for changed behavior. The appointment may need to change; supporting someone includes being willing to change the job rather than promise that more coaching will make it fit.

Before the next performance review, open the delivery plan. Find the work the new lead used to own. If their name has been replaced, is their time still required to get it finished?