Every care pilot begins with assumptions. Leaders assume referral partners will identify the right people, that staff can deliver the workflow reliably, that participants will accept the service, that data will be available quickly enough, and that the model will influence outcomes important enough to justify continuation. Yet many of these assumptions remain unwritten. Teams discuss them in design meetings, then move into delivery without a formal way to track whether they are proving true, becoming shaky, or needing to be replaced. Strong pilot evaluation and learning loops benefit from making those assumptions explicit through a structured log. For organizations developing new service models, an assumption log becomes a practical bridge between design thinking, live evidence, and governance.
In U.S. community services, this matters because pilots often fail for reasons that were visible from the beginning but never formalized. A county partner may not actually have the referral discipline leaders expected. A home-based model may demand more travel capacity than originally assumed. A target population may need stronger language support or more flexible engagement methods. Funders, commissioners, and boards increasingly expect providers to explain not just what happened in a pilot, but what core assumptions were tested and how the organization responded when some of them turned out to be wrong. An assumption log helps answer those questions. It turns hidden premises into governed learning rather than leaving them buried inside retrospective narrative.
Why implicit assumptions weaken pilot learning
When assumptions remain implicit, problems are often recognized late and explained poorly. Teams may say a pilot had “implementation challenges” or that “partner engagement was variable,” but without showing that the pilot was built on a specific assumption now proven unreliable. This weakens both learning and accountability. It becomes harder to know whether the model itself was poor, whether the context was misread, or whether the organization simply failed to test a critical dependency early enough.
Two oversight expectations support more discipline here. First, funders and system partners increasingly expect pilots to show how major design assumptions were tested rather than treating unexpected barriers as unforeseeable surprises. Second, boards, quality committees, and regulators usually expect leaders to demonstrate how identified risks and weak dependencies were escalated, reviewed, and acted on through governance. An assumption log supports both expectations by creating a visible record of what the organization believed at launch, what evidence later showed, and what response followed.
What an assumption log actually contains
A practical assumption log is not a theoretical exercise. It is a living record of important propositions underlying the pilot. Each entry usually includes the assumption itself, the part of the model it affects, the evidence that will test it, the owner responsible for review, the current status, and the action required if the assumption appears invalid or only partly true. Not every minor expectation needs to be listed. The log should focus on assumptions material enough to affect safety, access, workforce design, partner readiness, fidelity, cost, or scale potential.
Operational example 1: Logging referral assumptions in a community behavioral health pilot
What happens in day-to-day delivery
A community behavioral health navigation pilot starts with a formal assumption log reviewed every month at the implementation meeting. One key assumption reads: referral partners will identify eligible participants accurately enough that at least a defined proportion of referrals will convert to enrollment without major re-screening burden. The pilot manager assigns ownership to the intake lead and analyst. Over the first six weeks, the team tracks referral completeness, rejection reasons, time spent clarifying eligibility, and differences between referral sources. The assumption log is updated monthly with evidence summaries and a status field marked as holding, uncertain, or disproven. When one major referral stream consistently sends cases outside the agreed target population, the governance group records the assumption as disproven in its original form and agrees a revised referral training and triage process.
Why the practice exists and the failure mode it addresses
This practice exists because pilots frequently assume external partners understand the model’s eligibility logic more clearly than they actually do. The failure mode is allowing poor referral fit to be treated as routine operational noise rather than recognizing it as a failed design assumption. Without a log, teams can spend months firefighting intake issues without naming that the pilot’s access pathway rests on a premise that has now been proven unreliable.
What goes wrong if it is absent
Without an explicit assumption log, leaders may continue describing the pilot as fundamentally sound while blaming frontline staff for slow conversion or low enrollment. The real issue, however, may be that the referral design was based on an unrealistic view of partner behavior. That misunderstanding can distort the whole evaluation. Access appears weaker than expected, staff burden rises, and funders receive a vague story about implementation challenge rather than a clear analysis of which early assumption failed and what the organization did about it.
What observable outcome it produces
When referral assumptions are tracked explicitly, the organization can intervene earlier and explain changes more clearly. Observable outcomes include better targeted referral training, fewer inappropriate referrals, reduced intake rework, and a stronger governance trail showing that partner-dependent access risks were recognized and corrected rather than normalized. The pilot’s final report also becomes sharper because it can distinguish between the original assumption and the revised pathway that ultimately proved workable.
Assumption logs help separate model logic from context dependence
One of the greatest values of an assumption log is that it reveals which parts of the pilot are intrinsic to the model and which depend heavily on local context. This matters for scale. A pilot may appear successful, but only because a partner agency, staffing configuration, or data-sharing arrangement behaved exactly as hoped. If that dependency is made visible early, leaders can decide whether to redesign around it, formalize it, or limit future rollout to settings where the assumption is likely to hold.
Operational example 2: Testing staffing assumptions in a home-based maternal support pilot
What happens in day-to-day delivery
A home-based maternal support pilot includes a staffing assumption in its log from the outset: nurses and community health workers can jointly cover the target geography and visit intensity without creating excessive overtime, delay, or supervision burden. The workforce planner, service manager, and analyst review this assumption every month using route data, missed visits, overtime hours, escalation lag, and staff reflections on travel and scheduling pressure. Early evidence shows that the assumption holds in urban zones but is weak in rural catchment areas where visit clustering is harder and urgent follow-up disrupts routes more heavily than expected. The governance group does not simply note that staffing is “challenging.” It updates the assumption log to reflect that the original staffing premise holds only under certain geographic conditions and agrees revised deployment rules.
Why the practice exists and the failure mode it addresses
This practice exists because staffing models are often built on optimistic assumptions about geography, availability, and how much informal flexibility workers can sustain. The failure mode is believing the model is universally workable when, in fact, it depends on context-specific conditions that were never surfaced clearly. An assumption log helps reveal when a pilot’s apparent success is contingent rather than generalizable.
What goes wrong if it is absent
If the staffing assumption remains implicit, leaders may interpret delays and overtime as isolated operational teething problems rather than evidence that the deployment model itself needs revision. This can lead to poor scale decisions, unrealistic budgets, and unnecessary strain on staff. Participants then experience rescheduling, slower follow-up, or inconsistent coverage, while the final evaluation fails to explain that the real issue was not workforce commitment but a flawed starting assumption about what the staffing model could support.
What observable outcome it produces
When staffing assumptions are logged and reviewed, the organization can make earlier, more proportionate changes. Observable benefits include smarter route allocation, more realistic supervision ratios, clearer rural or urban eligibility boundaries, and stronger funding conversations because leaders can show exactly which conditions are required for the model to operate safely and reliably.
Assumption logs also improve reporting of mixed or changing results
Pilots rarely remain fixed. As evidence accumulates, some assumptions are confirmed, others are refined, and a few are retired entirely. A log gives reporting teams a disciplined way to explain this evolution. Instead of saying vaguely that the model “adapted over time,” they can describe which assumptions held, which proved too optimistic, and how those changes affected design, interpretation, and next steps. This improves trust because it shows learning as a governed process rather than a post-hoc story.
Operational example 3: Tracking participant-uptake assumptions in a respite and caregiver support pilot
What happens in day-to-day delivery
A caregiver support pilot begins with an assumption that families will take up a short-notice respite offer if the service is framed as flexible and low bureaucracy. The pilot director records this in the assumption log with evidence routes including inquiry-to-booking conversion, reasons for refusal, repeat-use patterns, and caregiver feedback on trust, preparedness, and disruption. Over two months, the evidence shows a more complex reality. Families value flexibility, but many do not accept short-notice offers unless continuity and pre-visit familiarity have already been established. The assumption is therefore revised rather than discarded. The governance team records that flexibility matters only when relational trust is in place, and the pilot changes its booking approach to include introductory contact before urgent use is expected.
Why the practice exists and the failure mode it addresses
This practice exists because participant uptake is often treated as a simple function of need or convenience when it is actually shaped by trust, timing, and perceived risk. The failure mode is assuming low uptake means weak demand or that strong demand will naturally translate into use. An assumption log helps the team identify the more precise condition under which the model becomes acceptable to families.
What goes wrong if it is absent
Without a formal record of the original uptake assumption, the organization may interpret weak early booking as a participant problem or a communication issue without realizing that its service design relied on an oversimplified premise. Leaders might keep pushing promotion or referral volume instead of redesigning how trust and familiarity are established. The result is slower learning, weaker caregiver experience, and a final report that describes adaptation without clearly showing what underlying belief had to change.
What observable outcome it produces
When uptake assumptions are tracked explicitly, redesign becomes more focused. Observable outcomes include higher booking conversion after trust-building steps are added, better alignment between caregiver feedback and service design, and stronger evidence for funders because the pilot can explain exactly what condition had to be met before demand translated into sustained use.
What leaders should ask when reviewing a pilot assumption log
Leaders should ask which assumptions are most material to safety, access, staffing, partner dependence, and scale; what evidence is being used to test them; which are now confirmed, weakened, or disproven; and what action follows when an assumption changes status. They should also expect the final evaluation to reference the most important assumptions and how they evolved. If that record does not exist, the pilot may have learned a great deal without being able to explain its own learning clearly.
The strongest pilot organizations do not just measure outcomes. They track the beliefs the model depends on and test those beliefs openly as delivery unfolds. That is what makes assumption logs valuable. They sharpen governance, surface hidden risk early, and improve the quality of both redesign and reporting. In practical terms, they help turn a pilot from a hopeful experiment into a disciplined test of whether a model can work under real-world conditions.