Workplaces
The great flattening: companies are cutting managers, not just jobs
In the first week of October, HubSpot, FICO and Amazon each cut roles with the same stated aim: fewer layers between the people who decide and the people who do the work. The cost savings are modest. The leadership consequences may not be.
At a glance: Between 6 and 8 October 2026, three US technology companies announced job cuts framed around organisational design rather than cost. HubSpot is removing about 660 roles, around 7% of staff, to build a flatter company. FICO told investors it will eliminate about 15% of positions, partly by "reducing the number of layers". Amazon removed a small number of roles in its Stores business. The common thread is middle management, and the risk is a thinner bench of future leaders.
Three companies, one design choice
On 6 October, HubSpot filed a Form 8-K with the US Securities and Exchange Commission (SEC) confirming that its board had authorised a restructuring plan on 1 October. The plan removes approximately 7% of the workforce and is expected to cost between $65 million and $75 million, mainly in severance, notice periods and transition support. The filing says the purpose is to "create a flatter, faster organization".
Chief executive Yamini Rangan was more direct in her memo to employees, which HubSpot filed alongside the 8-K. About 660 people are affected. "We will reduce management layers and move decisions closer to the people doing the work," she wrote. Product teams will be organised around customer outcomes rather than product hubs. She was also explicit about what the cuts are not: "This is not driven by AI-related efficiencies," and "This is not simply a cost-cutting exercise." The reduction, she wrote, "was the outcome of that work, not the starting point."
On the same day, Fair Isaac Corporation, the analytics company behind the FICO score, filed its own 8-K. Management committed on 1 October to a plan that "involves the elimination of approximately 15% of positions across the Company". The stated goals are "reducing the number of layers in the organization", "simplifying the operating structure", "optimizing processes and tools" and "integrating AI-driven product development". FICO expects about $27.0 million in pre-tax charges, mostly cash severance, and aims to complete the plan by the end of the third quarter of fiscal 2027. Affected employees began to be notified in the week of 5 October.
Two days later, Amazon confirmed it had made "the difficult decision to eliminate a small number of roles" in its Stores business, which runs its main retail website. The company did not give a number. Its statement said: "We're always looking at our team structures to ensure we're best set up to move fast as we innovate for customers."
The Amazon move is small, but it follows a long campaign. In October 2025, Amazon's senior vice president of People Experience and Technology, Beth Galetti, announced about 14,000 corporate role reductions and said the company wanted to be "organized more leanly, with fewer layers and more ownership". In January 2026 she announced about 16,000 more, describing the aim as "reducing layers, increasing ownership, and removing bureaucracy".
Why the middle is the target
Read together, the filings share a vocabulary. Layers, speed, ownership and decisions appear far more often than cost or margin. That is a deliberate signal.
For the past decade, management layers have been the default way to scale. A manager with eight direct reports needs a manager above them once the organisation passes a few hundred people, and each layer adds review steps, meetings and handoffs. In stable conditions this structure creates control. In conditions where products and pricing are being rebuilt around AI, as HubSpot and FICO both describe, leaders appear to see it as drag.
WCRC analysis: three forces are converging on the middle.
Tools now do part of the coordination job. Much of what middle managers have historically done, such as collating status updates, summarising performance data and routing information upward, can increasingly be handled by software. FICO names "optimizing processes and tools" and AI-driven product development alongside layer reduction. HubSpot says AI is not the reason for its cuts, but its strategy is built around delivering customer outcomes with AI, and its new structure assumes fewer people relaying information between teams.
Speed has become a stated strategic goal. HubSpot wants "a flatter, faster organization". Amazon wants to "move fast". When leaders believe the competitive window for AI products is short, each layer of approval looks like a cost measured in weeks.
Investors reward simplicity. HubSpot reaffirmed its third-quarter and full-year guidance alongside the announcement and said it remained on track for the long-term margin targets set out at its Analyst Day on 17 September. Framing a cut as redesign, rather than retreat, protects the growth story.
AI: reason, accelerant or narrative?
The three companies tell different stories about AI, and leaders should notice the differences.
HubSpot's chief executive went out of her way to say AI efficiency did not drive the cuts. FICO lists AI-driven product development as one of four aims. Amazon's statement on the Stores cuts did not mention AI at all, although Galetti's October 2025 memo called "this generation of AI" the most transformative technology since the internet.
The wider data shows why the framing matters. Outplacement firm Challenger, Gray & Christmas reports that artificial intelligence was the leading reason given for announced US job cuts in the first nine months of 2026, cited in 120,136 cuts, about 21% of the total. Companies that link layoffs to AI attract attention from employees, regulators and the public, and California has just passed a law requiring employers to state in mass layoff notices when AI or automation is a substantial cause.
WCRC analysis: a precise explanation is now a governance requirement, not a communications choice. HubSpot's decision to separate structure from AI, and FICO's decision to name both, are each defensible. A vague answer is not.
The leadership capacity risk
The case for flattening is clear. The cost is harder to see and arrives later.
Middle management is where most large companies grow their senior leaders. First-line and mid-level managers learn to hire, to deliver difficult feedback, to manage a budget and to make trade-offs under pressure. Remove a layer and the organisation removes many of the roles in which those skills are learnt. In three to five years, the bench for director and vice-president roles may be thinner.
There is also a load problem. When a layer goes, its work does not disappear. Spans of control widen, and the remaining managers absorb more direct reports, more one-to-ones and more people decisions. Without better tools and clearer decision rights, the result can be slower decisions, not faster ones, and burnout among the managers the company most wants to keep.
Finally, flatter structures change how people progress. In a company with fewer managerial steps, ambitious employees see fewer promotions. If the business does not build credible senior specialist tracks, with pay and status to match, it risks losing strong individual contributors who conclude that the only route up is out.
None of this argues against flattening. It argues for treating leadership capacity as something the redesign has to protect deliberately.
How the best redesigns are run
WCRC analysis: based on what HubSpot and FICO have disclosed, and on common practice in large reorganisations, several design choices separate a durable flattening from a one-off cut.
Start with decisions, not boxes. HubSpot's memo frames the cuts as the outcome of a design exercise. Leaders should be able to say which decisions move down a level, who now owns them and what information those people need.
Set a ceiling on spans of control. Wider spans only work if managers spend less time on reporting and more on coaching. That requires investment in tools and in removing low-value meetings.
Protect the pipeline. If layers come out, leadership development has to come from elsewhere: rotations, project leadership, stretch assignments and formal programmes for high-potential staff who no longer have a team to run.
Be exact about AI. If AI is a cause, say so and explain how. If it is not, say that too. Both HubSpot and FICO put their reasoning in SEC filings, which sets a standard others will be measured against.
Treat people well on the way out. HubSpot's memo sets out severance of 20 weeks of base pay plus one week per year of service, up to 30 weeks. How a company handles departures shapes how those who stay judge the new structure.
What it means for leaders
- Map decision rights before cutting layers. Know which decisions move, to whom, and what data they need to make them.
- Watch spans of control. Set a working limit and invest in tools that reduce reporting load before widening teams further.
- Rebuild the leadership pipeline on purpose. Create stretch roles, rotations and specialist career tracks to replace the development once provided by management layers.
- Explain the role of AI precisely and consistently across filings, memos and public statements, particularly in jurisdictions such as California.
- Measure the outcome. Track decision speed, manager workload and regretted attrition for at least a year after the change.
Quick answers
Q: How many jobs is HubSpot cutting in October 2026? A: About 660 roles, roughly 7% of its workforce, under a plan approved by its board on 1 October 2026. HubSpot expects charges of $65 million to $75 million and aims to finish by the end of the first quarter of fiscal 2027.
Q: What did FICO's 8-K say about its job cuts? A: FICO's 8-K, signed on 6 October 2026, says the company will eliminate approximately 15% of positions to reduce organisational layers, simplify its structure, improve processes and tools and integrate AI-driven product development, with about $27.0 million in pre-tax charges.
Q: Did HubSpot blame AI for its layoffs? A: No. Chief executive Yamini Rangan wrote that the cuts were "not driven by AI-related efficiencies" and were "not simply a cost-cutting exercise", but a move to a flatter structure with fewer management layers.
Filed under