Crestwood Group
Crestwood Group is a fictitious organization. This scenario is a composite drawn from patterns observed across privately held mid-market companies building data capability at the executive level for the first time. It is included here to illustrate strategic thinking and leadership approach.
Crestwood Group had grown primarily through acquisition. Over 15 years, the founding family had assembled three operationally distinct business units into a single holding structure. Each unit ran its own systems, tracked its own KPIs, and reported to the corporate executive team through monthly slide decks assembled by unit finance leads.
The executive team was experienced and commercially sharp. They were not data people. Decisions at the corporate level were made through a combination of financial summaries, gut instinct developed over years in the industry, and conversations with unit general managers who were themselves working from incomplete information.
This approach had worked well enough during a period of steady growth. It started showing its limits when the board began pushing for a more active portfolio management strategy, including the potential acquisition of a fourth business unit and the possible divestiture of underperforming assets within the existing three.
The CFO commissioned an external capability assessment. The conclusion was direct: the executive team did not have the data infrastructure or the analytical fluency to support the level of portfolio decision-making the board was now expecting.
The capability assessment had identified the problem clearly enough, but identifying a problem and fixing it are different challenges. Two previous attempts to improve reporting at Crestwood had failed. The first produced a dashboard that nobody used. The second produced a data warehouse project that ran over budget and was quietly shelved.
The executive team's relationship with data initiatives was cautious at best. The CFO was aware of this and framed the engagement accordingly. The goal was not to build another dashboard. It was to change how the executive team thought about and used data in their day-to-day decision-making, and to build the infrastructure that would support that shift.
Four problems were driving the gap between the data the business generated and the decisions the executive team was making.
First, nobody agreed on what the business actually looked like in numbers. Each unit tracked revenue, margin, and utilization differently. Corporate rollups required manual reconciliation and still produced figures that unit GMs would dispute in the monthly review meeting.
Second, the executive team had no shared language for data. The CEO and CFO used the same terms to mean different things, and neither realized it until decisions were already made on conflicting assumptions.
Third, data arrived too late to influence decisions. The monthly reporting cycle meant that by the time the executive team saw performance data, the relevant decisions had already been made operationally. The data confirmed what had happened. It did not shape what would happen.
Fourth, the unit GMs experienced data requests from corporate as a burden rather than a benefit. Every request required manual extraction and formatting. There was no shared infrastructure, no automation, and no clear line between the effort of producing data and the value it created for the people producing it.
The first thing established was a decision inventory. Rather than asking what data the company had, the starting question was: what are the ten most important decisions this executive team makes in a given year, and what information would make each of those decisions better?
That exercise took three sessions with the CEO and CFO. It produced a concrete list that became the design brief for everything that followed. The data infrastructure, the dashboards, the training, all of it was anchored to those ten decisions. If a data point did not connect to one of them, it did not get built.
This approach addressed something the previous two initiatives had missed. They had been built around what data was available. This one was built around what decisions needed to be made.
Before any platform was built, a definitions workshop was run with the executive team and the three unit GMs. The goal was to agree on a single definition for twelve core business metrics: revenue, gross margin, utilization rate, client retention, overhead ratio, and seven others specific to Crestwood's portfolio.
The workshop took two full days and surfaced disagreements that had been generating friction in monthly reviews for years. The revenue recognition difference between the real estate unit and the facilities management unit alone had been producing a $4M discrepancy in the corporate rollup that nobody had formally resolved.
Getting to agreed definitions before building anything was slower than most technology projects would allow. It was also the reason the platform that followed was actually used.
The platform built for Crestwood was deliberately limited in scope. Three dashboards, one for each business unit, feeding into a single corporate portfolio view. Each dashboard showed the metrics connected to the decision inventory, updated weekly from unit finance systems through an automated extraction process.
Visual design decisions were made with the executive audience in mind. No charts that required explanation. No metric that lacked a plain-language definition on hover. A traffic light system for performance against target that could be read in under 30 seconds. The CFO reviewed every dashboard before release and flagged anything that required interpretation. Those elements were redesigned until they did not.
Access to better data is not the same as the ability to use it. A parallel program ran alongside the platform build to develop the executive team's data fluency directly.
Six sessions were run over four months, one per month with the full executive team and one per month with the unit GMs separately. The sessions were not training in the traditional sense. They were structured around real business questions using real Crestwood data. Each session started with a decision the executive team was actually facing, worked through what the data said, and ended with a documented view on what additional information would have made the decision sharper.
The CEO was the most engaged participant in those sessions. He had been privately uncomfortable with his own data fluency for years. Having a structured space to develop it, without that discomfort being visible to the broader organization, made a significant difference to his engagement and to his willingness to use the platform publicly with the board.
Months 1 and 2: Foundation
Months 3 and 5: Build
Months 6 and 8: Adopt
Months 9 and 10: Sustain
| Metric | Target |
|---|---|
| Core business metrics with agreed definitions | 12 of 12 before platform build begins |
| Corporate rollup discrepancy resolved | Yes, before first quarterly review |
| Monthly manual slide deck process retired | Month 7 |
| Executive team data fluency sessions completed | 6 sessions over 4 months |
| Board portfolio review delivered from platform data | Month 10 |
| Unit GM satisfaction with corporate data requests | Measurable improvement vs. baseline |
| Phase 2 analytics roadmap documented | Month 10 |
The definitions workshop in month one surfaced the $4M revenue recognition discrepancy that had been sitting unresolved in the corporate rollup for three years. Resolving it before the platform was built meant the numbers the executive team saw from day one were numbers everyone agreed on. That had never been true before.
The CEO used the corporate portfolio dashboard in a board meeting for the first time in month eight. He walked the board through three months of performance data without notes, without a slide deck prepared by the finance team, and without deferring to the CFO for interpretation. The board chair commented afterward that it was the most confident presentation of business performance he had seen from the executive team.
The unit GMs' relationship with corporate data requests changed once the automated extraction process removed the manual burden. By month nine, two of the three GMs had asked for additional views within their unit dashboards. They had gone from experiencing data as something done to them to something they could use.
The CFO, who had watched two previous data initiatives fail, described the difference plainly. The previous initiatives had started with technology. This one started with what decisions needed to be made and worked backward. That sequencing was what made it land.
Crestwood entered the next fiscal year with an executive team capable of conducting the portfolio-level analysis the board had been asking for. The potential acquisition of a fourth business unit, which had stalled in part because the executive team lacked the analytical infrastructure to evaluate it rigorously, moved back onto the active agenda.
Crestwood Group is a fictitious organization. This scenario is a composite drawn from patterns observed across privately held mid-market companies building data capability at the executive level for the first time. It is included here to illustrate strategic thinking and leadership approach.