The Owner's assumed operating model holds in two dimensions and fails in four. Where the audit confirms the assumed model: brand-standard adherence in front-of-house technical execution; consistency of service in the morning daypart. Where the audit contradicts the assumed model: staff do not believe leadership knows what their day actually looks like; recovery authority lives on the org chart but not on the floor; the “hospitality first” brand promise is read by staff as “hospitality when nothing else is breaking”; and the turnover that leadership attributes to wage competition is, on the staff side, attributed to a single supervisor.
The audit finds six structural patterns from staff lived experience that the Owner did not know were operating: three confirm the assumed model (and should be preserved through current renovation planning), three contradict it (and require leadership attention before they compound into Q3 turnover spikes and guest-review erosion).
The most consequential finding is structural and personnel-adjacent: the supervisor on the 3 p.m.–11 p.m. shift in housekeeping is the single largest predictor of voluntary separation in the property's 14-month tenure under current ownership. Twelve of the seventeen separations in that period were direct reports of this supervisor. This is not surfaced in exit interviews because the interviews are conducted by the supervisor's direct manager. The audit recommends the Owner take this finding offline before any other action — and route the personnel decision via labor counsel.
Recommended next actions route three findings to sister Hospitality agents (HOS-04 care diagnostic, HOS-03 standards system, HOS-05 service recovery), one finding to the Talent suite (TAL-04 performance-imperative diversity architect, given the demographic concentration in the affected shift), and two findings to the Managing Partner's direct decision (the supervisor-pattern decision; the floor-authority threshold for recovery).
What this deliverable covers: Reading the staff lived experience of working at the property — what staff actually do in a shift, what they think the assumed operating model is, where their lived experience confirms or contradicts it, and the structural patterns that result. Five culture dimensions are assessed (Authority, Recognition, Recovery, Belonging, Voice). Each finding is evidence-vs-inference distinguished; severity-rated; routed to the right next-step owner.
What this deliverable explicitly does not cover: Individual performance reviews; HR investigations; compensation analysis; the remedy build (standards system, recovery playbook, accommodation architecture, anticipation loops — those are sister-agent work). Findings that touch a single named employee's tenure are flagged for the Managing Partner's direct attention and routed offline; they are not detailed in the body of this audit.
The audit reads the staff lived experience through five inputs, triangulated:
The audit does not use anonymous all-staff surveys. The signal-to-noise on those is too poor, and they invite the kind of language staff already know leadership will not act on. Twenty-three structured conversations under confidentiality discipline produce more honest signal than 248 anonymous surveys.
The audit rates each dimension on whether the staff lived experience confirms or contradicts the assumed operating model. Confirm means leadership's mental model and the floor reality are aligned. Contradict means they are not, and the gap is being absorbed by staff at a cost.
| Dimension | Definition | Status | Severity |
|---|---|---|---|
| Authority | Do staff have the authority on the floor that the org chart says they have? | Contradicts | High |
| Recognition | Does the property see the work staff actually do, or only the work the brand standard measures? | Partially contradicts | Moderate |
| Recovery | When a guest issue surfaces, can the staff person in front of the guest resolve it? | Contradicts | Structural |
| Belonging | Do staff experience the property as a place that wants them long-term? | Contradicts within a single shift cohort | High — concentrated |
| Voice | Do staff have a channel to surface what they know about the property that leadership does not? | Confirms partially | Moderate |
Pattern. Authority on the floor lags authority on the org chart by twelve recurring scenarios. Staff cope by either making the call anyway (and absorbing the risk personally) or by deferring — which the guest reads as “the person in front of me cannot help me.”
Routing. Owner-level decision on the recovery-authority threshold (see Finding 4); standards system rebuild routes to HOS-03.
Pattern. Recognition systems — bonuses, employee-of-the-month, brand-award nominations — are calibrated to the brand-standard number. Staff who carry the extra workload are not visible to the recognition system. The recognition system is functionally invisible to the people doing the work the property most depends on.
Routing. Recognition-architecture rebuild routes to HOS-03 (standards system) with a calibrated workload model; Talent advisory on retention-impact routes to TAL-04.
Pattern. The recovery architecture exists on paper. On the floor, it functionally does not. Staff resolve what they can on personal initiative, defer what they cannot, and the unrecovered moments compound into the guest-review erosion the property is currently experiencing.
Routing. Service Recovery Playbook routes to HOS-05; the empowerment-threshold design needs to be set by the Owner (recovery dollar limits, comp authority by role, decision rights without escalation) before HOS-05 builds the operational system.
Pattern. The turnover story leadership has told itself (“wage competition”) is not what the data show. The data show a concentrated personnel pattern that the property's feedback architecture is structurally unable to surface. This is a Belonging-dimension failure expressed as a turnover statistic, and it is also a structural diversity question because the affected shift cohort is demographically homogeneous in a way the rest of the property is not.
Routing. Owner-direct decision. Personnel action and any associated process must route via labor counsel, not via this firm. The audit names the pattern; the action is yours. Talent advisory on the demographic concentration and downstream retention architecture routes to TAL-04 after the personnel question is resolved.
Pattern. The property has no “staff-heard” intake channel. Staff hear early signal that, if logged and synthesized, would reach leadership before it reaches the next guest's review. Today it stays on the floor and dies there.
Routing. Care diagnostic crosses to HOS-04 for guest-side cross-check on each named pattern; ADA pattern routes to ADA-04 (Compliance & Public-Accommodation Audit) for risk assessment before it becomes a Title III demand-letter exposure.
Pattern. One of the property's genuine strengths is at structural risk: the 7:15 walk-through is informal, ungoverned, and dependent on two specific people. The Q4 renovation planning currently proposes restructuring engineering's morning coverage in a way that would functionally end the walk-through.
Routing. Anticipation Framework (HOS-01) to formalize the walk-through into a structural practice that survives the renovation; standards system (HOS-03) to embed it in property SOP.
Read together, the six findings describe one underlying pattern: the property's feedback architecture is structurally biased toward confirming the assumed operating model rather than surfacing where it diverges from the floor reality.
The brand audits measure technical execution against standard. The engagement surveys are anonymous and conducted at intervals long enough that staff edit. The exit interviews are conducted by the supervisor whose behavior is most often the reason for separation. The Manager-on-Duty system exists in name but not on the floor. The recognition system rewards adherence to a workload model that no longer matches the actual workload.
Each of these is a small structural choice. Together they produce a property that genuinely does not know what its own staff lived experience is — and a leadership team that, in good faith, is acting on a mental model the floor has quietly stopped matching.
This is repairable. None of the six findings is a referendum on individual leaders. Five of the six are repairable through architecture changes the firm's sister Hospitality agents are built to make. The sixth is a personnel decision that belongs to the Owner and labor counsel.
| Priority | Action | Routes to | Timeline |
|---|---|---|---|
| 1 | Take Finding 4 (supervisor / shift-cohort pattern) offline. Engage labor counsel. Do not act on it inside the property until counsel has the file. | Managing Partner direct · labor counsel | This week |
| 2 | Set the recovery-empowerment thresholds (dollar limits, comp authority, decision rights) before any other recovery work begins. | Managing Partner direct, then HOS-05 builds against the thresholds | 2 weeks |
| 3 | Build the service recovery playbook against the new thresholds. | HOS-05 | 4–6 weeks |
| 4 | Rebuild the recognition system to a calibrated-workload model (Finding 2). | HOS-03 · cross-check with TAL-04 | 6–8 weeks |
| 5 | Cross-check the three named guest-sentiment patterns (Finding 5) against guest verbatims and against ADA risk exposure on the accessibility-request pattern. | HOS-04 · ADA-04 | 4 weeks |
| 6 | Formalize the morning-shift walk-through (Finding 6) and protect it through the Q4 renovation planning. | HOS-01 · HOS-03 | Pre-renovation; 4 weeks |
| 7 | Re-audit at 12 months. The structural patterns named here should be measurably different in the lived experience; if they are not, the architecture changes did not land. | HOS-06 (this agent) | 12 months |
Sources: twenty-three confidential staff conversations across departments and shifts (off-property, anonymized at synthesis); ten shadow shifts (two per department across rotations); the property's SOP set, training matrix, accident/incident log, exit-interview database, 24 months of engagement-survey raw data, and brand-standard compliance audits; structured assumed-model conversations with the Managing Partner, GM, Director of Operations, and Director of HR; the property's 12 months of guest verbatims for cross-check. All client-confidential sources are described categorically. No external guest, employee, or competitor name appears in this brief.
Method: lived-experience audits read what staff actually experience, not what surveys produce. The five-dimension framework (Authority, Recognition, Recovery, Belonging, Voice) is the firm's standard culture-audit instrument; it is calibrated per engagement to the property segment, brand standard, and ownership context. The deliverable is structural, not personal — with the explicit exception of Finding 4, which the audit names because the data require it and which the Owner must act on with labor counsel, not inside the property.
This brief is delivered as the firm's senior advisory analysis to the Managing Partner. The AI agent assembled the staff-conversation synthesis and the assumed-model triangulation; my judgment shaped the dignity-standard framing, the sequencing of recommended actions, and the explicit decision to route Finding 4 offline to labor counsel rather than operationalize it inside this brief. The brief does not leave the firm until I sign it. Not legal advice on employment or labor matters — those must be routed to counsel.