How should managers be rewarded? At Microsoft, and many other companies, rewards supposedly are based on business and customer impact. That sounds great, but doesn’t that approach skew rewards toward managers disproportionately? After all, individual contributors (ICs) are judged only on their own impact, while managers are judged on the impact of everyone reporting to them. Heck, even when a team has a poor performer, the manager can get rewarded for removing that person.
Anyone who’s been in people discussions knows that managers tend to receive higher rewards than ICs receive. This trend often gets rationalized by claiming that managers are chosen from the best and brightest, but that’s absurd. The best and the brightest don’t always make good managers, and all people within the same compensation band are expected to be capable at their respective roles. Nonetheless, managers tend to be rewarded at a higher rate due to their perceived higher impact. The way economists deal with this kind of disparity is to normalize impact by an appropriate denominator. (I’ll also talk about impact numerators compatible with current people discussions.)
The obvious denominator is the count of staff. Individual contributors would divide impact by 1, and managers would divide impact by their org size (including themselves). However, if the teams of two different managers produced roughly the same impact, but the first manager’s team was all senior engineers and the second’s team was all entry-level ICs, then shouldn’t the second manager’s impact count more? Likewise, an entry-level IC with comparable impact to a senior IC should be valued more (or the senior should be valued less). Thus, the proper denominator is compensation budget (org or individual). You’d be surprised how incentives would change if we normalized impact this way.
Compensation budget is provided by Microsoft’s review tool today. It includes salary, bonus, and stock. Unfamiliar with people discussions? Read Out of calibration.
David and Goliath
Normalizing impact by compensation budget would remove the disparity between managers and ICs, but that’s not all. Today, managers of big orgs have a big advantage over peer managers with small orgs, even if the small orgs produce more impact per person. I’ve seen multiple examples of managers receiving the same rewards as peers whose teams produced comparable results with half the staff and a third of the compensation budget. The org leaders sometimes even acknowledge the difference in org size, but claim that the managers produced the same impact, as if org size and budget were irrelevant to running a business.
For example, say Dick’s group of 40 engineers produces $30M of impact during a review period, while Jane’s group of 20 engineers produces $24M (using dollars for impact in this basic example). Dick’s impact was 25% more than Jane’s, so today he’d likely receive higher rewards. Say Dick’s staff had roughly the same compensation budget per person as Jane’s ($300K per person, fully burdened). Dick’s compensation budget ($12M) is twice Jane’s ($6M), making Dick’s normalized impact 2.5 and Jane’s normalized impact 4. Jane’s normalized impact was 60% more than Dick’s—her team was more impactful per person and per dollar spent. Still think Dick deserves higher rewards?
If we normalize impact by compensation budget, large teams must produce correspondingly more impact to compare favorably to small teams. Since small teams typically are more responsive and productive than large ones (as I describe in Staying small), normalizing impact would likely kill empire building and headcount hoarding, resulting in faster market response, more innovative teams, and lower operating expenses.
Normalizing impact does ignore the additional difficulty of running a 1,000-person org versus a 10-person org. However, reward systems account for scope through compensation bands and their associated levels. Normalized impact should be used to compare the impact of peers within a compensation band, or perhaps across bands, when a promotion or demotion is under consideration.
To learn the perfect team size for productivity and impact, read Span sanity—ideal feature teams.
No growth for you
Today, there’s a desire to bring in young diverse talent and grow them into leaders, but there’s little incentive for managers to do so. Yes, there’s incentive to bring in young diverse talent—they are easy to source and hire thanks to outstanding high school and college recruiting programs. However, that talent ends up leaving after a few years. Why? Because there’s no room to grow into the leadership ranks. Why? Because a team made up of principal engineers is likely to have a greater impact than the same-sized team made up of senior and entry-level engineers. Since budgets are somehow irrelevant to rewards, managers today are incentivized to create and maintain top-heavy teams that have little room for senior and entry-level people to advance.
Budget concerns can pressure executives to change the mix of levels within their orgs. However, the tools available to executives that trim headcount budget are crude: attrition, layoffs, hiring freezes, and promotion freezes. Those tools work short-term but are painful and don’t fix the underlying problem.
If we normalize impact by compensation budget instead, top-heavy teams will compare unfavorably to balanced teams. Every team has a mix of work, but top-heavy teams will pay more for it (as they literally do financially). Top-heavy teams will also fail to benefit from the above-level impact of lower-level engineers. This will likely induce managers to spread principal engineers across groups rather than concentrating them on one team. This approach further incentivizes level-appropriate work and creates more room for engineers to advance. Creating room is particularly important to underrepresented engineers who benefit from diverse role models and mentors. Taken together, this method of normalizing impact generates greater value for the same or less expense.
Corrupt managers might withhold promotions, but that’s an unstable strategy. Engineers already leave teams when overlooked for promotions. With an incentive to hire under-leveled talent and greater opportunity for under-leveled engineers to advance, corrupt managers won’t keep teams long. To avoid the issue entirely, companies can track managers’ promotion metrics.
Normalizing impact by compensation budget will cause trouble for high-level engineers who don’t consistently deliver commensurate impact. That’s rough but appropriate. Potential remedies include seeking broader impact, switching to a more motivating role, or lowering their level.
What’s it worth to you?
If compensation budget is the divisor for normalized impact, what’s the numerator? Executives responsible for P&Ls measure impact in revenue, profit, and/or engagement. Unfortunately, the impact of directors, group managers, leads, and ICs is not as easily ascribed. Creative leaders might apportion revenue, profit, or engagement among their organizations, but that approach is fraught with subjective assignment, double-counting, and missing collaborative contributions.
A simpler approach is to assign impact values during org planning, then adjust those values based on results. Top priorities would have the highest initial impact values, while other efforts are proportionally lower. At review time, leaders accumulate and adjust impact values for work completed, including contributions across teams and results built upon the work of others. Leaders apply extra credit for exceeding goals and apply debits for falling short, including potential penalties for outages, people issues, or other concerns. The total impact values are divided by compensation budget to reach a normalized impact for the review period.
While normalizing impact will solve many problems, assigning a specific measure to impact may overlook key factors, such as organizational health (including growth, inclusion, and retention), corporate principles and values, and other contributions that don’t align to planned work. Therefore, robust people discussions are still necessary to consider all aspects of each employee’s contribution. Just remember to consider compensation budget when comparing impact.
At Microsoft, people discussions often use cards that list key information for people, such as their level and length in level. Compensation budget should be listed next to level since both raise expectations when comparing relative impact. For more details, read Discussing rewards for people.
For more on budgeting, read On budget and It’s business time. For more on growth and inclusion, read Growth mindset and diversity.
It ain’t right
It’s wrong to give managers higher rewards, on average, than those given to ICs within the same band. It’s wrong to give managers of large teams higher rewards, on average, than those given to managers of small teams within the same band, ignoring productivity and responsiveness. It’s wrong to reward managers for hoarding headcount or high-level talent. It’s wrong to reward managers for blocking young diverse talent from growing into leadership. It’s wrong, and it needs to change.
By normalizing impact by compensation budget, we can remove the disparity between managers and ICs, kill empire building and headcount hoarding, create room for young diverse talent to grow, gain faster market response and greater innovation, and lower operating expenses. It’s a simple correction that companies have somehow missed for decades. It’s time to right some wrongs.
Want personalized coaching on this topic or any other challenge? Schedule a free, confidential call. I provide one-on-one career coaching with an emphasis on underrepresented, midcareer software professionals. Find out more at Ally for Onlys in Tech.
The idea is clear.
What about measuring the impact? If one team develops Software and another team – Hardware, and then company sells both together as a package? And what about teams that do internal products?
Management is typically responsible for evaluating impact and allocating rewards. Management can specify the planned impact values for the software and hardware teams in advance, helping those team prioritize. Once those teams deliver their pieces, management can adjust the impact values based on work completed, with extra credit for exceeding goals and penalties for outages, people issues, or other concerns. Management can then divide final impact values by compensation budget to reach a normalized impact for the review period.