Two retreats. Six teams. One heartbeat.

Not a turnaround. A profitable, respected organization with recent industry awards, a technical team already deep in AI-assisted development, and a CEO who could name the problem out loud: decisions felt top-down, and the work felt fragmented. Thirty people in nine time zones will do that. Every agreement has to survive a twelve-hour gap, and most of them didn’t.

Agile had been installed five years earlier and had done its job: the ship was out of port. But the prescription had been applied to every team the same way, and the teams whose work isn’t software — communications, partnerships, support, data — were force-fitting their world into sprints, complexity points and story cards that were never built for it. Sixty-eight initiatives sat on the product backlog. Planning was weak, work in progress was everywhere, and the certifications in the room were not the problem.

The ask, in the leadership’s words: team agile training and certification, facilitate the strategic-planning day, harden the org chart, and get a handle on the work in progress. The brief, said in front of the whole company: real coaching and human connection, not another framework.

We didn’t open with a workshop. We opened as a fly on the wall: six weeks of cycle planning, executive roadmapping, sprint planning, backlog refinement, retrospectives and sprint reviews across every team, with the camera on and the mouth mostly shut. The pattern was the same at every altitude.

One disease, three symptoms: agreements existed on paper; nothing enforced them; a few patient people were the compensating mechanism. And the part that mattered most: the organization had already said this about itself. We didn’t need to sell the diagnosis. We needed to help them act on it, in their own vocabulary, without adding to anyone’s load. The company had just published its values, drafted from staff words, and two of them were almost literally the fix: planning that feels expected rather than confrontational, and measuring what’s said against what’s done. Every move we made was retagged in those words. No Scrum sermon, no Kanban sermon, no transformation program.

In June, before the engagement had paper, we offered the leadership team four formats for their one in-person day of the year: a World Café, a multi-team agile exercise, a hands-on AI workshop, and a talk. They took all four and built the day around them. By the time the planning committee’s near-final agenda arrived, four of the day’s blocks had our name on them:

One Frequency is nine steps and a break. No slides to open with; lights down and two slow breaths instead. The mission read aloud, slowly, twice, with silence between — not analyzed, felt. One honest word from everyone, anonymously, posted and clustered live on the wall. The positive cluster is the guiding star: what “good” feels like here. The negative cluster is data, not complaints: each one points at the exact place the mission and the lived week don’t match yet. Then each team writes the promises that protect the good side of the wall, and names the constraints in its way.

The week around it mattered as much as the day: the staff-and-board social the night before, one-on-ones from the executive team down to the newest analyst, and, two days later, a written retro to the CEO: agile is installed and the ship is out of port; the “how” had been more prescriptive than each team needed; the working agreements need real paper and signatures to come alive; and the two non-software teams needed breathing room to understand the why before being handed another how. The to-do list that followed was short and specific: time with the team leads alone, and a version of agile each team could call its own.

Two weeks out, the CEO shared the strategic-planning deck and his speaker notes and asked for “any insider nuance.” We reviewed the final deck with the executive team the afternoon before and wrote our own facilitator notes: the arc of the day, the questions to hold, where the pivot was, which table carried the most signal, and the one insight to use sparingly, once, well. Posture: light. His state of the union; our structure, our facilitation, our futurecasting.

Nearly thirty people in the room — board, the stakeholder associations that own the company, and staff. The morning opened with the CEO’s core question: four co-equal objectives, or one north star and three enabling constraints? Then a World Café we facilitated: five tables, one per objective plus a wildcard for new priorities and emerging risks, groups rotating, every table leaving its thinking behind for the next, convergence harvested at the end. The same discipline as the all-hands: we notice and name; we don’t author.

The afternoon was board and executives only, and it was ours: futurecasting. Six questions about the next five years — which products to spin out or keep, whether to pool data with peers, what the ownership and board structure should be, how to license what the company owns. A temperature-check vote across all six, the deep dive on the one with the most energy, and the other five dated as agenda items rather than dropped. The frame we set: these aren’t decisions today. They’re the questions that decide 2031. Where the question touched ownership, we held it strictly neutral: facilitate, never opine.

Two questions we carried in reserve and asked once, neutrally: if these three objectives are constraints, which one is the constraint right now? And what’s the one number that tells us we’re growing, not just getting leaner? The next morning the executive team met to digest the day with us in the room. A week later the CEO’s first draft of the 2027 strategic direction came back to us for suggestions, feeding the October board and the November budget. One board-room attendee wrote the next day to ask for the same session for her own team.

Five days after the all-hands, the whole executive team received a written operating model. The thesis: agile is installed; the “why” behind the “what” is missing in places; each team now needs a cadence it truly chooses itself, with real KPIs both the team and leadership agree to, matched to one core heartbeat of leadership planning. Remove complexity: move each team’s work from the unknowable to the documented to the repeatable. Feelings became working agreements (how we do together) became constraint identification (how we work with others). Now KPIs per cycle, per team, on whatever cadence lets that team come up for air and align as a whole.

A month later, after the fly-on-the-wall review of every team meeting and its permutations was complete, the plan for the last quarter put the heartbeat to work as a funnel, up and down:

The AI-enablement thread ran underneath all of it. The CEO’s own product planning was already coalescing into a durable sequence with an AI assistant, and the engineering team was deep in AI-assisted development. The operating model was built for that world: five years of entrenched agile, decomposed for the age of AI — fewer complexity points, less story-card ceremony, more flow, a heartbeat that gets instrumented rather than administered, and hiring and roles considered for an AI-enabled company. Every sprint is just an experiment, iterating until consistent delivery is normal and the margin for growth, learning and science increases.

Ten days after the strategy day, the CEO who had sponsored the work moved on. The posture changed the same afternoon: stability and continuity — listen, steady, stay small, take no side. The team meetings continued. The interim leadership, a finance chief stepping up and the longest-tenured executive as operating lead, got a short written note rather than a proposal: take the time to find the right CEO, no rush; the gap is the executives’ chance to set the new reality and raise their own authority; a quarterly planning cadence with the engineering lead’s rule of one project planned, one executing, one wrapping up is exactly right for the constraints; and we can help on two fronts, the organization’s structure, cadence and heartbeat with leadership, and cadence, expectations and alignment with the teams.

At the October retrospective with both agile leads, the teams had already transcribed their own all-hands boards into a written synthesis, and it landed where our July notes had: the commitments between colleagues were real; what stood in the way was structural — priorities, intake, capacity. Working rules were set for the next phase: we work only with team leads who can decide; leads source their own teams; the framework starts from the boards the company already wrote. Interim leadership then chose to close the engagement in November. Because every artifact was the teams’ own, nothing left with anyone.

No lift is claimed on this page. We didn’t baseline a metric, so we won’t invent one. What follows is what exists, and who owns it.