One of the most overlooked barriers to successful scale is not the new model itself, but the old one that never fully leaves. Providers may launch a stronger pathway, open new sites, train staff, and communicate revised referral criteria, yet still keep legacy processes running in parallel out of caution, habit, or stakeholder pressure. As explored across the Impact Insights Hub’s work on scaling what works and its broader analysis of new service models, scale becomes fragile when old and new models coexist without clear boundaries. Duplicate referral routes remain open, outdated forms are still used, supervisors keep reviewing cases against prior expectations, and local teams quietly preserve earlier ways of working “just in case.” The result is confusion, duplicated effort, blurred accountability, and reduced confidence in which model is actually being delivered. Scaling what works therefore depends not only on implementation, but on disciplined retirement of the pathways, assumptions, and workflows the new model is replacing.
Why legacy systems continue to undermine scaled models
Community services rarely begin on a blank page. New models are usually layered onto pre-existing teams, referral habits, partner expectations, documentation templates, and governance arrangements. This means expansion does not only involve building something new. It also involves unbuilding parts of what came before. Providers often underestimate how difficult that is. Staff may retain old spreadsheets, partners may keep sending referrals through historic routes, and managers may continue reporting against legacy measures because they feel more familiar or politically safer.
This matters because running old and new systems in parallel creates operational drag. Staff waste time reconciling duplicate information, referrers become uncertain about where cases should go, performance data becomes harder to interpret, and the new model struggles to demonstrate its real value because it is carrying the cost and confusion of the old one. Unless legacy retirement is treated as a core scaling task, expansion can look active while the service remains structurally split between past and future.
What a credible legacy-retirement approach should include
A credible approach should identify which previous routes, forms, roles, metrics, and governance habits are being replaced, what the transition timeline is, and who has authority to close the old system rather than merely de-prioritize it. It should also define short-term exceptions carefully so that temporary overlap does not become permanent duplication. Most importantly, it should communicate clearly to frontline staff and partners what has ended, what remains, and where accountability now sits.
Strong providers also audit whether retirement has really happened. It is not enough to announce a new pathway if case flow, documentation, and supervision continue to reflect the old one. Retirement has to be visible in practice, not just in strategy papers.
Operational example 1: Closing duplicate referral routes in a scaled discharge-support model
In day-to-day delivery, a hospital-to-home stabilization model replaces a patchwork of earlier referral routes, including ward-specific email requests, local spreadsheets, and informal phone escalation to individual coordinators. During scale-up, the provider does not simply launch a new centralized intake form and hope referrers switch over naturally. It creates a defined retirement plan that closes old mailboxes, redirects informal requests into the new pathway, retrains discharge teams, and monitors which routes are still being used by mistake. For a short transitional period, any legacy-route referral is accepted only if it is re-entered immediately into the new intake architecture and the referring team is corrected on the same day.
This practice exists because one common failure mode in scaling is leaving legacy access routes open for convenience. While this may feel helpful during transition, it usually prolongs confusion and prevents the new model from stabilizing. Staff continue checking multiple channels, referrers never fully learn the new system, and leaders lose confidence in the integrity of the front door. The retirement plan exists to create one accountable route rather than an informal network of exceptions.
If this control is absent, the operational consequence includes duplicated work, slower triage, and weaker cohort control. Some referrals enter the new system, others arrive through old channels, and staff spend time reconciling information instead of managing risk and follow-up. Over time, the provider may believe the new front door is underperforming, when the real problem is that it has never been allowed to become the sole route of entry. This undermines both fidelity and evaluation.
The observable outcome includes cleaner referral behavior, faster intake visibility, better data quality, and stronger confidence that the scaled model is operating through one coherent access architecture. It also reduces hidden administrative burden and helps new sites adopt the model more consistently because they are not inheriting parallel intake expectations from older practice.
Operational example 2: Retiring legacy review habits in a behavioral-health continuity model
In routine delivery, a behavioral-health continuity service introduces a new model built around structured continuity-risk review, shared escalation definitions, and clearer thresholds for urgent re-engagement. However, some supervisors continue using older case-review habits shaped by the pre-scale service, where continuity was managed more informally and decisions were less standardized. To address this, the provider identifies legacy supervisory practices that conflict with the new model, replaces outdated review templates, recalibrates supervisors together, and audits whether case discussions are being conducted against the new framework rather than prior local custom.
This practice exists because another major failure mode in scaling is the survival of old professional habits beneath new documentation. Staff may appear to be using the new pathway, but if supervisors are still reinforcing earlier threshold culture, local interpretation will remain inconsistent. Legacy retirement therefore has to include review behavior, not just forms and technology. Otherwise the old model continues to shape decisions from underneath.
If this function is absent, the operational consequence includes hidden model conflict. Frontline teams receive mixed messages about what counts as continuity risk, what should trigger escalation, and how quickly follow-up needs to happen. Supervisors may think they are being pragmatic, but in practice they are preserving site-specific legacy standards that prevent the new model from becoming genuinely shared. This weakens fidelity and makes cross-site comparison difficult.
The observable outcome includes more consistent supervision, clearer threshold culture, stronger staff understanding of what has actually changed, and better protection against drift masquerading as professional discretion. It also helps the service demonstrate that scale has produced a new common operating method rather than just a new label applied to older practice.
Operational example 3: Decommissioning outdated reporting and oversight structures in a multi-partner network
In day-to-day practice, a lead provider scales a community support model across several local partners while replacing an older service arrangement that used locally varied spreadsheets, monthly narrative summaries, and different performance definitions in each county. To prevent the old reporting environment from continuing indefinitely, the provider works with commissioners to retire legacy measures, close duplicate reporting templates, and align all partners to one dashboard and one assurance rhythm. During the transition, the provider tracks where legacy reporting is still being requested and escalates these issues quickly so the old framework does not continue by default.
This practice exists because a further common scaling failure mode is legacy reporting persistence. Even when the new model has better measures and clearer quality logic, older reporting requirements often survive because they are familiar to local commissioners or embedded in past contract habits. This forces staff and managers to serve two accountability systems at once, weakening the new model’s ability to define its own performance clearly. Decommissioning old oversight structures is therefore essential if the new model is to become governable at scale.
If this process is absent, the operational consequence includes administrative overload, confused accountability, and an inability to interpret performance cleanly. Partners may focus on whichever reporting regime matters most locally, creating inconsistent behavior across the network. The new model then appears more cumbersome than it really is because it is still carrying the weight of the system it was supposed to replace.
The observable outcome includes cleaner reporting discipline, stronger partner alignment, better dashboard integrity, and more credible evidence that the scaled model is being governed through one set of performance expectations. This gives both providers and commissioners a much clearer basis for improvement, comparison, and future expansion.
Commissioner and oversight expectations
Commissioners increasingly expect providers to explain what old pathways or processes are being replaced when a service is scaled, not just what new features are being introduced. They want to know how duplicate routes will be closed, what transitional overlap is unavoidable, and how long that overlap will last. Funders also benefit from this clarity because it reduces the risk of paying for a new model that never fully displaces the inefficiencies of the old one.
Oversight bodies generally look for consistency between stated service design and actual operating practice. Providers should be able to show that outdated access routes, documentation habits, and review structures have been retired in practice rather than simply de-emphasized. Where overlap remains temporarily necessary, they should be able to explain how it is controlled and when it will end.
Why this matters now
As more community services move from pilot or redesign into full-scale implementation, legacy pathway retirement is becoming a decisive part of whether the new model ever achieves its intended value. Services that fail to retire the old system often blame the new model for confusion, delay, or duplication that actually comes from unresolved coexistence. Services that decommission legacy routes and habits deliberately are more likely to achieve cleaner scale, stronger accountability, and better operational confidence. In practical terms, scaling what works depends not only on launching the new, but on having the discipline to close what no longer should govern the work.