In February 2024, Klarna announced that an OpenAI-powered assistant was doing the work of 700 customer service agents in its first month, projecting a $40M profit improvement. By December, the company had cut headcount from 5,527 to roughly 3,422, a 22% reduction. Then Klarna reversed course and started rehiring humans, with the CEO admitting the company "overestimated AI's capabilities."
The press releases overstate the pace of knowledge-worker elimination. The structural shift underneath the headlines is that AI is collapsing the optimal team size.
Companies are running the math on what 100 employees used to do, and figuring out how to do it with 10. What 1,000 used to do, and figuring out how to do it with 100. The teams getting smaller are getting more leveraged, and the per-person cost of any single bad interaction is getting catastrophic. Someone at every Fortune 500 company is going to own that problem by 2028. They will have a title we have not standardized yet, a budget that did not exist five years ago, and a mandate carved out of what used to be the CHRO's job.
Call them what you want. Director of Team Dynamics. Head of Team Effectiveness. Chief Squad Officer. Head of Org Effectiveness. The label is converging.
The shrinkage is real, and uneven
Marc Benioff said in February 2025 that Salesforce was seriously debating not hiring any engineers in 2025, citing 30%+ productivity gains from Agentforce. Microsoft has done more than 15,000 layoffs across two rounds in 2025, and Satya Nadella has called the company's 220,000+ headcount a "massive disadvantage" in the AI race. Amazon cut 14,000 corporate jobs in October 2025, with Andy Jassy explicitly saying "we will need fewer people doing some of the jobs that are being done today." Tobi Lütke sent a memo at Shopify in April 2025 instructing managers to demonstrate "why they cannot get what they want done using AI" before requesting headcount.
The trend is uneven. Anthropic is at 5,000 employees and growing. WEF's 2025 Future of Jobs report projects 78 million new jobs by 2030 alongside the disruption. McKinsey's State of AI 2025 found 32% of companies expect AI to reduce total workforce by at least 3% in the next year, and WEF says 41% of employers plan to reduce workforce as AI automates tasks. Both numbers are real. So is the rehiring.
The shape of the change is what matters for how companies organize people. Reid Hoffman put it cleanly in Superagency: humans will do "high-leverage work by directing teams of agents, so that a single person operates with the capacity of an entire team". Sam Altman said the one-person billion-dollar company is now possible. Whether that exact threshold gets hit, the trajectory is small teams doing large-team work, with each individual node carrying more weight.
The economic case for tiny teams
Tomasz Tunguz's "communication tax" piece lays out the math. A 150-person traditional org has 11,175 communication channels, by Brooks's classic formula. A 30-person AI-enabled org has 435. That is a 96% reduction in coordination overhead. The same piece points out that an Anthropic engineer named Boris Cherny ships 10 to 30 PRs per day, where the industry baseline is closer to 3 per week. Roughly 30x productivity per person.
Anthropic does roughly $5M in revenue per employee with around 5,000 people, where Apple does $2.4M per employee. Cursor (Anysphere) hit $2B ARR by April 2026 with somewhere between 150 and 300 employees, putting them around $6.7M per head. These are leading indicators of a broader pattern in how AI-leveraged orgs deliver output.
Once you accept that small teams can do large-team work, the next question is what makes a small team work. Richard Hackman's research at Harvard, summarized in his canonical "Five Keys to Successful Teams" work, found that optimal team size is 4 to 6 and never more than 10, with communication problems increasing exponentially as team size grows. Brooks's Law captures the same point in software. Bezos invented the two-pizza rule for the same reason.
The math punishes friction. In a 50-person department, one bad interaction is noise. In a 5-person AI-leveraged squad doing the work of 50, one bad interaction is a quarter going sideways. The cost of conflict, dysfunction, miscommunication, and trust breakdown scales inversely with team size. Small squads with high cohesion compound. Small squads fighting each other stall.
The CHRO function was built for a different era
Look at what is happening to the people who currently own this. Russell Reynolds tracked 155 CHRO appointments globally in 2025, up 25% year over year. HR Brew reported 30 new CHRO/CPO appointments at Fortune 200 firms in 2024, a 36% YoY increase in turnover. The job is the second hardest in the C-suite according to Fortune 200 board chairs. Gallup measured US engagement at 31% in 2024, a 10-year low. Global engagement is at 21%. Gartner reports HR functions spend $2,810 per employee annually on tech, and 51% of HR leaders cannot measure ROI on those investments.
None of that is a personal failure. The role itself was built for broadcast culture. Annual engagement surveys, company-wide values rollouts, all-hands programming, big benefits design, leadership academies that scale to thousands. When the company was 50 departments of 100 people each, broadcast worked. The CHRO's job was to set the tone and let it propagate.
When the company is 50 squads of 5, the broadcast loses signal. A 5-person squad with three tenured experts and two new hires has its own operating culture. Whatever the CHRO sends from the top filters through three layers and lands as background noise. The Q12 engagement question that has declined the most over the past decade is "I know what is expected of me." That question lives at the squad-operating layer, well below where culture comms can reach.
The function will not disappear. The Gartner 2025 survey found leader and manager development is the #1 HR priority for the third consecutive year. There is plenty of work for the CHRO. The function is being reorganized, and a new role is splitting off underneath it.
What the Director of Team Dynamics actually owns
The function we are watching emerge is squad-level operating performance, measured in behavior, owned by someone with budget. Specifically:
- Decision velocity inside small teams. How fast can a 5-person squad get from disagreement to commitment? In a tiny org, this is the bottleneck.
- Role complementarity. Who in this squad is the closer? Who is the connector? Who is the explorer? Putting the wrong five people in a room kills the project. QuestWorks calls these HeroTypes; other systems use different language. The work is the same.
- Conflict recovery time. When two of five people fight, the squad is dead until they reconcile. The Director of Team Dynamics owns the protocol for fast recovery.
- Onboarding into squads with established rhythms. Adding a sixth person to a 5-person squad triggers a category change in how the group operates. Time-to-trust is a real metric.
- Org Network Analysis. Deloitte estimates 50% of collaboration happens outside formal reporting lines. ONA tools detect burnout in central nodes 3.5x faster than surveys. The new role owns that data layer.
- Psychological safety practice, at the squad level. Google's Project Aristotle studied 180 teams and found psychological safety was the #1 dynamic separating effective from ineffective teams. Every CHRO has cited this for a decade. Almost none have figured out how to operationalize it past a survey question. Squad-level work is where it gets practiced.
This is the $359 billion problem CPP estimated in 2008, recast for the era when teams are small enough that friction shows up in the weekly numbers. Gartner reports 74% of HR leaders saw increased disputes after RTO, and 72% of organizations have no formal conflict resolution policy. The need for an owner is acute.
Customer Success, RevOps, People Ops: this is what category emergence looks like
None of this is unprecedented. The pattern is:
- The work exists, fragmented across a half-dozen titles.
- A vendor or thought leader brands the category.
- The title crystallizes.
- It is mainstream within five to ten years.
Customer Success ran this playbook in the open. From 2000 through 2005, the work was scattered across Account Manager, Client Services, Customer Care, and Implementation roles. Gainsight launched in 2009 and branded the category. By 2015, "Chief Customer Officer" was a Fortune 500 title. The function had existed across all those years. The title arrived later.
Revenue Operations followed the same arc, faster. Through 2018, Sales Ops and Marketing Ops and CS Ops were three separate teams that fought each other. Clari, Gong, and Outreach branded "Revenue Intelligence" and "RevOps." The Chief Revenue Officer and VP RevOps roles were standard within five years.
People Operations is the closest analogue to what is happening now. Laszlo Bock rebranded HR as "People Operations" at Google in 2006. Project Aristotle ran from 2012 to 2015 and made people analytics a real discipline. By the late 2010s, "People Ops" was a normal company function. None of those titles existed in 2005.
The Director of Team Dynamics role is in step 1 to 2 right now. The work is fragmented across agile coaches, OD consultants, learning and development, ONA platform admins, people analytics leads, and external team-coaching firms. i4cp's 2025 People Analytics Leaders survey shows titles like "Head of Organizational Effectiveness and Workforce Analytics" and "Head of Workforce Planning and Insights" already exist at major companies. Half of people analytics leaders see GenAI as important or very important to their function in 2025. Spotify's squad model (Squads of 6 to 12, Tribes of up to 100, Chapters, Guilds) operationalizes some of this.
The branding moment has not happened yet. When it does, the title will crystallize fast. By 2028, every Fortune 500 will have someone in the seat.
The boundaries of the role
The CHRO still owns total rewards, compliance, succession, executive coaching, and the broad organizational health story. The Director of Team Dynamics sits below that, with a tighter scope: a budget-holding, function-owning role focused on the operating performance of small teams, with hard metrics. Think of it as the operating layer between the CHRO's strategic culture mandate and the actual work. The CHRO owns the company's culture story. The Director of Team Dynamics owns whether each squad ships.
The role is also distinct from a coach-in-residence, an ONA admin, and an L&D rebrand. Each of those is one input. The Director of Team Dynamics consolidates them into a single function with a single owner.
The shrinkage story is uneven. Some companies will scale up. Some functions will grow. The role exists wherever teams are small enough that per-person friction shows up in the operating numbers, which is increasingly everywhere knowledge work happens.
What this means for buyers
If you run people strategy at a Fortune 500 today, the practical move is to start consolidating fragmented work under a single director before the market makes the role obvious. Agile coaches, OD consultants, ONA platform owners, team-coaching contractors, squad health surveys, the manager development budget that Gartner says is the top HR priority. Pull the thread together.
Companies that built RevOps in 2018 ate the companies that built it in 2022. Companies that built People Operations in 2010 had a four-year window where the function was a competitive advantage, before it became table stakes. The Team Dynamics role is in the same window now.
The metrics that matter for this role are squad-level operating signals: decision velocity, time-to-trust on new hires, conflict recovery time, role-complementarity scores, repeat-collaboration rates from ONA. Most of these are not standard data sources yet, which is part of why the role has not crystallized. The data layer is being built right now by a few different categories of vendor.
The flight simulator for team dynamics
The new function needs a substrate. Engagement surveys measure individual sentiment at the company level, well above where a 5-person squad operates. ONA gets directional signal but does not change behavior on its own. Performance reviews are too lagged.
What the role needs is data from teams actually working together, in repeatable conditions. We built QuestWorks as the flight simulator for team dynamics. It runs 25-minute weekly cinematic quests for groups of 2 to 5 on its own voice-controlled platform. The platform generates 9 HeroType archetypes from each player's behavior, surfaces them publicly via QuestDash with behavioral callouts visible to all players, and produces a separate weekly team health report for leaders showing aggregate trends. HeroGPT lives in Slack and provides private 1:1 coaching that never shares conversations upstream. Voluntary, opt-in, never tied to performance reviews. $20 per user per month, 14-day free trial. Works with Slack.
QuestWorks is one tool a Director of Team Dynamics might use, alongside ONA platforms, team-coach networks, and whatever new vendors emerge in the next two years. The role needs telemetry from teams in motion, and the category of products that provides it is just starting to mature. No single tool defines the seat.
2028 is not far
Customer Success went from "what does that title even mean?" to a Fortune 500 standard in 15 years. RevOps did it in 5. People Operations did it in about 10. The Director of Team Dynamics role is in the same starting position those roles were in their year-one phase.
The pressure is higher this time. The CHRO churn is at 25% YoY. Engagement is at a 10-year low. Companies are doing 10x with a tenth of the people. The work that gets done in 5-person AI-leveraged squads has too much per-person leverage to leave squad dynamics to chance.
By 2028, every Fortune 500 will have someone with budget authority owning team dynamics. The smart move is to build the role before the market forces it.
Try it free
