The 90-day stabilization phase after an EHR go-live is not a calendar window. It is the stabilization boundary — the signed exit point where the named FDE operator transfers the program to a named client-side owner, running without us on the floor. We pick up the engagement at go-live. We exit on three signs, not three months: adoption measured, integration stable, the named internal owner running it.
The first adoption signal we read is helpdesk ticket shape — not volume. In week one, ticket categories are dominated by access and configuration: password resets, role assignments, the bartender who was missed in the new module roster. That is expected. By weeks four to eight, the category mix shifts. Tickets are about workflow drift — the order sets that did not survive the floor, the handoff paths the floor does not actually run, the documentation written for the design cohort and not for the running cohort. Volume alone is a poor signal because the volume we want to see rise early is the volume of calls the floor feels safe making. The category mix tells you whether the floor is asking the right questions or whether the readiness list is drifting under the queue.
The second adoption signal is super-user utilization, and we measure it against floor reach rather than logins. A super-user program is only as strong as the named people in it and the floors they actually walk. A super-user who has logged in forty times this month but has not been on the floor in three is a super-user we are about to lose. A super-user whose visits to three named floors drop from weekly to monthly is a near-miss, not a usage report. We read utilization against cell six — same Friday, same named client-side owner — and a fall in utilization is the cell getting quieter before it goes dark. Names, not log files.
The third adoption signal is near-miss reporting, and the direction of the trend matters more than the number. The frontline is supposed to capture close-call workflow failures — the order entry that almost went to the wrong patient, the medication reconciliation that nearly missed a flagged interaction, the discharge summary that was about to leave with a stale allergy on it — before they are incidents. Near-miss volume up, across categories, is a positive signal: the floor is trusting the reporting channel, names are being written down, the readiness list is being read against actual floor behavior rather than against an idealized one. Near-miss volume down with the same throughput is the readiness list drifting silently. The channel we opened is still being read; the floor has stopped writing into it. That is a stabilization signal, not a usage pattern, and we treat it the same way we treat super-user falloff: name the cell, name the owner, walk it on the next Friday.
Our named FDE operator is still on the floor through this window. Not on a ticket-queue rotation, not on a remote status report. The same operator who carried go-live is still carrying the Friday review against cell six, still named, still embedded next to the named clinical and operational owners on your side. They are not closing tickets — they are reading the readiness list against your floor so the named internal owner can take over the read on the day we exit. If the named FDE operator who carried your cutover is not the same named operator answering the page in month two, the stabilization phase has already drifted. We do not let it drift. Operator continuity is one of the cells we hold hardest through the window.
The handoff back to internal IT closes when three things are true at the same time. The named client-side owner is running the program without us — no FDE on the Friday review, the named owner reads cell six on their own, the named clinical owners rotate through ownership rather than rotating out. The adoption measure holds six consecutive weeks against baseline, read by the same named owner, against the same six categories the readiness list was signed against in week zero. The Stabilization & Adoption Retainer check has been signed — the readiness list re-baselined, the cell-by-cell ledger updated, the named super-user rotation owned inside. When those three signs are signed, the named FDE operator transfers the program on the same week. When even one of them is not signed, we stay.
The handoff side of the same window — what the named operator side hands back on the day of exit, and what the CIOs we work with sign on that day — is also where the frequently asked questions on the FAQ page live. The CIO and CMIO question we hear on day sixty is the same one buyers should be asking on the other side of that window, and the buyer-side checklist on vendor deployment track record vetting walks through that same question against the eligibility roster before the contract goes out. The stabilization phase is the exit; the same checklist is the entry you read against before you sign.
Write the handoff into the readiness list on day one. Do not wait for month three to find out whether the program can run without us.