Wayne Rosing’s Problem
The Undoing of Middle Management
It’s finally here — the elimination of middle managers. It’s been a long time coming. Middle management’s existence is under threat. However, it doesn’t seem like middle managers are in an existential crisis, they have no doubts whatsoever regarding the relevance of their roles, purpose, and importance within organizations.
You have heard of the 2022 layoffs, followed by the 2023 layoffs, and now the 2024 layoffs. Who is being laid off? It's mostly middle management. 2022 layoffs were described as a normal rationalization of the workforce after tech had overhired during the pandemic boom. 2023 layoffs were explained away as the pursuit of efficiency for companies that were now prioritizing profitability over growth. Zuck called it “The Year of Efficiency” at Facebook. What shall we call 2024 layoffs? Some say it's the AI-driven layoffs — the reduced number of people needed in some roles as the wave of GenAI takes over.
All these sorts of layoffs, and their varying explanations are primarily eliminating layers of middle management, along with some baseline bottom-of-the-layer employees.
During the pandemic, it was almost obvious that middle managers would be targeted in the post-pandemic era due to the fast adoption of technology, and the history of managerial evolution. I was decidedly on the side of “killing the middle managers” and collected my thoughts in the article below, penned in August 2021.
Now that companies are doing exactly as predicted, I am starting to see cracks and faultlines in that line of reasoning.
Let's look at what happening out there:
Citi Group — plans to cut 20,000 jobs in 2024 in a massive delayering exercise meant to reduce the number of management layers from 13 to 9. It is basically middle management that is being targeted, facing the bullets.
Deloitte’s ambitious, restructuring plan meant to reduce complexity by shrinking main business units from 5 to 4, cutting costs across the firm, and “free up more workers to work with clients rather than manage staff internally”. Of course, this will involve delayering but when put across in corporate speak you might not even notice. Deloitte is a master of using the grandiose corporate vernacular. In general, in ‘corporate’ you don't say “We are firing middle managers”. You say “We are implementing a strategic realignment aimed at enhancing operational agility and maximizing client-centric focus, optimizing our organizational structure by rationalizing business units, streamlining processes, and empowering our workforce to deliver unparalleled value to our clients”
NVIDIA — Jensen Huang’s first principles approach to management led him to structure an organization in a manner that requires very few middle managers. He has 40 direct reports. That right there cleared 3 or 4 middle managers. The move also required complimentary changes like having a radical transparency system in place, so “everybody knows everything”, with no strategy secrets residing at the top.
Google had a funny experience in July 2001. Larry Page tried to fire all Product Managers, on the basis that Google did not need them. It didn’t go well. Google had 6 product managers at that time, now it has more than 2,000.
Larry and Sergey founded Google at a young age. Larry nursed the company through a rapid growth phase moving from having all employees directly reporting to him to having product managers sliding into existence between him and engineers. Larry fired all PMs and required all engineers to report to only one person, a very capable VP of Engineering named Wayne Rosing.
This reorganization was reversed quickly. Engineers were unhappy, some were now unproductive, some needed feedback on projects and their careers, and some were tired of dealing with issues they thought somebody else higher-up was supposed to deal with like mobilizing project resources, budgeting, communicating with externals and internals beyond your team, etc. So, Google started hiring Product Managers again.
Larry Page’s experiment was done 23 years ago. Things have surely evolved, and surely companies can now survive with fewer Product Managers. That is precisely the current zeitgeist. However, companies have to be careful, even though times have changed, there is still a limitation on what Wayne Rosing is capable of doing.
Wayne Rosing’s Problem
Now, if you were Wayne Rosing on that day, you would love to be a VP, but you are essentially presented with a huge problem, and that is managing all sorts of engineers at scale. This is the Wayne Rosing Problem.
Wayne Rosing has a problem. The engineers have a problem. The organization has a Wayne Rosing “bottleneck” problem whereby Wayne Rosing becomes the weakest link, the weakest point in the chain. The weight moves from 6 PMs to one Wayne Rosing. The dexterity of Wayne in effectively managing such a large team is not present in most people. Jensen Huang is an outlier.
When you suddenly get rid of a managerial layer, the managerial duties (yeah, the ones you cannot easily measure and monitor) do not disappear. The people who were shouldering those duties and responsibilities vanish, but the duties remain. Somebody has to pick them up. It is a burden for those who are left. The excessive delayerings that we are witnessing now could potentially create the Wayne Rosing Problem where things get worse instead of getting better.
The surviving workers are overworked, their direct reports are under-supervised, unsupported, unmentored, and demotivated. Though the goal was to gain some efficiencies, the unintended consequence is the opposite.
Sooner or later, Wayne Rosing will be requesting additional staff, Assistant VP of Engineering, Assistant to VP of Engineering, Director of Engineering, Chief Engineer, Director of Engineering projects — Automation, etc to help get stuff done. The bloat resurfaces, only that this time the job titles are not “Project Manager”.
Tech has enabled us to widen the spans of control. AI is now assisting in workflows so things have gotten way better over the past two decades vs the time Wayne Rosing had a problem. Reporting is in the process of being fully automated. The question is what do middle managers do?
The answer is that we might not monitor and measure their contributions visibly but they do contribute some value.
Is this the end of Middle Management? Maybe! Maybe Not!
The delayering — getting rid of middle management is one phenomenon. The other phenomenon is getting rid of low-level and entry-level jobs. It seems there is nowhere to run or hide except going upwards.
That is it!
Ciao!
