Reading Notes

Heidi Helfand on Dynamic Reteaming, the Five Patterns of Team Change, and Transparency in Reorgs

Source: Heidi Helfand on Dynamic Reteaming, the Five Patterns of Team Change, and Transparency in Reorgs

Notes — Heidi Helfand on Dynamic Reteaming, the Five Patterns of Team Change, and Transparency in Reorgs

Four questions [Adler frame]

Q1 — What is it about? A practitioner’s framework for understanding and managing team change at software organisations, presenting five recurrent patterns of “reteaming” — the inevitable evolution of team composition — as a replacement vocabulary for the overloaded, top-down concept of “reorg.”

Q2 — How is it argued? Through twenty years of personal experience at fast-growing SaaS startups (Expertcity/GoToMyPC, AppFolio), supplemented by research interviews with practitioners at Spotify, Procore, Redgate Software, and Menlo Innovations. Supporting frameworks borrowed from William Bridges (Transitions), RIDE (Pat Wadors), and co-active coaching.

Q3 — Is it true? The five patterns are descriptively accurate and provide a useful vocabulary. The prescriptive advice for the isolation pattern (process freedom, senior sponsorship, knowledge redundancy as precondition) is well-grounded in named examples. The anti-patterns are backed by concrete negative experiments. The normative claim — that involving people in reorg planning through whiteboard transparency is worth doing — is asserted more than rigorously tested, though the Procore and Spotify examples are plausible.

Q4 — What of it? Gives teams a shared vocabulary for something universally experienced but rarely discussed systematically. Most actionable: the isolation pattern checklist for spinning up an innovation team; the whiteboard reteaming approach for low-cost transparency; the switching pattern’s role in building knowledge redundancy before a crisis demands it.

Glossary

  • Dynamic reteaming: the ongoing, inevitable changing of team composition treated as a natural and manageable aspect of growing or shrinking software organisations; contrasted with the “stable teams” orthodoxy of Tuckman’s forming-storming-norming-performing model
  • One by one: people joining or leaving the company (distinct from #switching, which is moving between teams within the same company)
  • Grow and split: teams that grow beyond effective coordination size split into two or more focused teams; signals include longer meetings, decision slowdowns, divergent work, and inattention in standups
  • Merging: two or more teams or companies combine; the reverse of grow and split; typically a shrinking or consolidation pattern
  • Isolation / innovation by isolation: a new team placed off to the side with process freedom to catalyse a new product line or solve an emergency; a “beneficial silo” that runs against the common de-silo instinct
  • Switching: moving from one team to another within the same company for learning, fulfilment, or knowledge-redundancy purposes; a lateral transfer, not a company exit
  • Whiteboard reteaming: transparent org-change planning where proposed team structures, names, missions, and open slots are presented to the affected team on whiteboards for input before finalisation
  • RIDE framework: Request, Input, Decider, Execute — a decision-clarity tool for reorgs attributed to Pat Wadors (CPO Procore/UKG); distinguishes who requests the change, who can give input, who decides, and who executes
  • Percentage anti-pattern: allocating people as fractional resources across multiple concurrent efforts (10% here, 20% there); rarely adds up in practice and imposes heavy context-switching costs
  • Knowledge redundancy: ensuring multiple people share ownership of any given system through pairing and switching; the precondition that enables safe reteaming without stranding single-owner knowledge silos
  • Panarchy: the nested hierarchy of organisational levels (individual → team → team of teams → department → company) across which reteaming changes propagate and interact

§ The five patterns of reteaming

Heidi names five patterns and distinguishes each from “reorg” — a word she considers too loaded with top-down, involuntary connotations. Reteaming includes everything from Sue switching to the web ops team to learn something new, through to two companies merging.

One by one. Someone joins or leaves the company. Tip: help new joiners feel belonging through pairing on day one (not sitting alone). Equally important: coach existing team members through changes they weren’t involved in — for example, a new leader hired from outside who becomes their manager.

Grow and split. Teams grow beyond effective coordination size and divide. Signals: longer meetings, harder decisions, divergent work, standups where people have checked out. Problem-trading warning: splitting may create resource gaps (one PM, one designer across two teams). [?] Heidi does not give a specific team-size trigger; signals are behavioural.

Merging. Opposite of grow and split — two or more teams combine. Can be org-level (companies acquire each other) or team-level (attrition-driven consolidation). Tactic: “story of our team” (chapter 13, second edition) — each team makes a timeline of milestones and achievements, shares it, then builds a shared forward vision together.

Isolation / innovation by isolation. Full pattern and checklist at #§ Isolation pattern — implementation checklist.

Switching. Lateral team-to-team movement for learning, fulfilment, and knowledge redundancy. Extends employee lifespan at the company; creates safety nets against single-owner system risk. See #§ Switching and knowledge redundancy.

§ Isolation pattern — implementation checklist

The canonical example: Expertcity had built a marketplace for tech support that wasn’t selling (“made six bucks last month”). The CEO killed it and isolated a small team from the waterfall drag. That team, freed from pixel-perfect Photoshop mockups and heavyweight process, pivoted to GoToMyPC — a product that saved the company. Later expanded to GoToMeeting and GoToWebinar. Replicated at AppFolio with SecureDocs, which eventually spun out and was acquired in 2022.

Checklist:

  1. Separate location (at minimum, a distinct claimed region in the workspace — SecureDocs had their own corner with named walls)
  2. Tell other teams explicitly: do not disturb this team; a senior leader must voice this
  3. Build knowledge redundancy first (pairing, shared ownership on the old system) so the isolated person can fade out without leaving an unmaintainable single-owner mess
  4. Grant process freedom — let the team iterate at whatever cadence fits the problem; do not force two-week sprints onto one-day loops
  5. Report up to a senior decision-maker with real authority — decisions must not get reversed via bureaucracy; isolation fails if the team has to seek approval through a complex web
  6. Relieve bureaucracy — strip heavyweight reporting (quarterly slide decks, etc.)

Historic validation: the Chicken McNugget was saved by an isolated SWAT team at McDonald’s — worked in a different plant, reported directly to a senior executive (documented in Teamwork, 1970s).

§ Whiteboard reteaming and transparency

Inspired by Christian Lima at Spotify, who visualised a large infrastructure reteam on whiteboards and invited team input. Heidi replicated this at Procore with ~80 people in a platform organisation splitting from two large clusters into three.

Whiteboard contents: future team structure with everyone’s names in proposed positions, team names, team missions, and open hiring slots.

Outcomes observed: people identify design mistakes the leaders had missed; people spot internal opportunities they may apply for; agency and visibility create a sense of ownership rather than imposition.

Spectrum of transparency: whiteboard reteaming is mid-spectrum. Redgate Software (Chris Smith, Cambridge) runs open self-selection events where teams pitch their missions and people choose — more open, appropriate when the organisation is ready for it. The closed back-room model remains common because leaders fear distraction and conflict.

Precaution: time-box the input phase; bias towards shorter windows; the process should be expedient, not a weeks-long deliberation.

§ Anti-patterns

Percentage anti-pattern. Allocating people fractionally across multiple concurrent efforts (10% on A, 20% on B, 30% on C) — a Waterfall inheritance. Percentages do not add up in practice; cognitive context-switching costs are high; “full-stack” team ownership breaks down.

“Poof, they’re gone.” Changes with no communication — new hires arrive without announcement, people depart silently. The absence of communication removes the opportunity for the team to process transitions (Bridges: the neutral zone requires acknowledgement to navigate).

Spreading high performers. Jon Walker at AppFolio ran the experiment: identified a high-performing team with palpable chemistry and energy, then spread its members across other teams to import the magic. It did not work. Team chemistry is an emergent property of people, time, conditions, and working context — not a transferable resource that moves with individuals.

§ Switching and knowledge redundancy

Menlo Innovations (Richard Sheridan, Ann Arbor) built company-wide pair-switching into the founding model: all roles work in pairs and rotate at a regular cadence. This parity starts at interviewing — candidates experience pairing before they join.

Safety net framing: when companies downsize, no single owner of a critical system is lost. AppFolio’s ACH-processing systems required careful switching (not haphazard) because real-money flows were at risk — but a deliberate designed cadence gave team members agency and redundancy simultaneously.

Engineers on an AppFolio team that grew and split voluntarily started rotating between the two resulting teams to maintain working relationships — a self-organised switching pattern driven by fulfilment. Autonomy to do so was itself a motivating signal (cf. Dan Pink: autonomy, mastery, purpose).

§ Transitions framework (Bridges) and RIDE

Bridges’ Transitions (Transitions: Making Sense of Life’s Changes): Endings → Neutral Zone → New Beginning. The neutral zone is liminality — you are not yet comfortable in the new reality but are no longer in the old one. Leaders can raise positivity by painting the vision and purpose of the new state. Knowing this arc reduces anxiety in subsequent changes because the neutral zone becomes recognisable rather than alarming.

RIDE framework (Pat Wadors, CPO Procore/UKG): Request (who raised the change need), Input (who can contribute ideas), Decider (who makes the final call), Execute (who carries it out). Clarifying D in particular prevents the friction of ambiguous authority in complex reorgs.

Scope boundaries matter: not all changes admit input. “You’re getting acquired — you’re not part of that decision.” Clarity on RIDE scope prevents frustration when a change is outside employees’ agency. The team-level equivalent — “should we split into two?” — may legitimately involve the team in the D role.