Retention & Key Talent Plan — ACME Health / Cumberland Valley Integration

Which sixteen roles the program cannot afford to lose before the transition services agreement is exited, why those and not others, and what the retention instruments are actually buying. Issued September 5, 2023 — three weeks before closing, so that offers can be made in the closing week rather than after people have started drawing their own conclusions.

Retention does not buy loyalty, engagement or goodwill. It buys time, and specifically it buys the time to transfer knowledge — not the time to keep doing the job. That distinction decides everything else in this plan. If the purpose were continuity of work, the instruments would cover the largest number of people for the longest period. Because the purpose is transfer, they cover a small number of people whose knowledge is undocumented, for exactly as long as the transfer takes, and each one is paired with a named receiver who has to be able to do the job at the end of it. ⚠ A retention payment with no transfer obligation attached buys a person who stays and then leaves with everything still in their head.

Part I — Who and Why

Objectives
Identification Criteria
The Sixteen Covered Roles
The Asymmetry

Part II — The Instrument

Instrument Design
What Retention Does Not Buy
Knowledge Transfer Linkage

Part III — Running It

Communication
Measurement
When Someone Leaves Anyway

1. Objectives

Objective	Measure	Target
Critical knowledge available through TSA exit	Covered roles retained to their release date	⚠ Above 85%. Not 100% — a plan assuming nobody leaves is not a plan.
Knowledge transferred, not merely retained	Transfer accepted by reverse shadowing	Every covered role has an accepted transfer before release
No single point of knowledge at TSA exit	Roles with no trained receiver	Zero
People treated straightforwardly	⚠ Departures citing being misled	Zero. See §8.
Cost proportionate	Retention spend against the exposure it protects	Carried by Human Resources, not by the program budget

2. Identification Criteria

A role is covered if losing it before its release date would delay the TSA exit or create an unrecoverable knowledge loss. Seniority is not a criterion.

Ref	Criterion	Test applied
RC-01	Undocumented system knowledge	⚠ Can another person configure, reconcile or explain this system from documentation alone? If not, the role qualifies.
RC-02	Sole holder of a relationship	Provider, regulator or trading partner relationships not held elsewhere
RC-03	Named in a regulatory filing or license	Roles the regulator recognizes cannot be vacated without notification
RC-04	Required for TSA exit certification	The receiving side of a knowledge transfer that has not yet happened
RC-05	Clinical program leadership	⚠ The care management capability is preserved because it performs. The platform without the people is not the capability.
RC-06	Marketability	⚠ A modifier, not a criterion on its own. Applied because the people most worth keeping are the most able to leave.

RC-01 is the criterion that produces uncomfortable results, and the discomfort is the signal. It surfaces mid-level configuration analysts and managers rather than executives — because the person who can explain why a claims edit exists is rarely the person who runs the department. A retention list built by seniority protects the org chart; a retention list built by RC-01 protects the program, and the two overlap far less than anyone expects. ⚠ Where a senior role appears on this list it is because of RC-02 or RC-03, not because of its level.

3. The Sixteen Covered Roles

Tier A is released at TSA exit; Tier B at the completion of its workstream; Tier C at Day 100.

Tier	Individual	Role	Criterion and release
A	J. Whitfield	Configuration Analyst, Claims — retiring platform	⚠ RC-01. The claims edit inventory depends on this role. Released at core admin cutover + 60 days.
A	C. Bhattacharya	Manager, Member Data	⚠ RC-01. Single point of knowledge for the target's member data structures. Released at TSA exit.
A	D. Okpara	Manager, Claims Configuration	RC-01, RC-04. Released at core admin cutover + 60 days.
A	J. Amonett	Manager, Interfaces	RC-01. Write-back and accumulator interfaces. Released at cutover.
A	G. Whitmire	Director, IT Operations	RC-01, RC-04. Data center and TSA exit certification. Released at TSA exit.
A	Dr. P. Nwachukwu	VP Care Management	⚠ RC-05. Preserved capability — release is indefinite, not dated.
A	A. Boudreaux	Director, Provider Contracting	RC-02. Provider relationships and contract history. Released at network rationalization completion.
A	D. Halloran	Business Analyst — enrollment & eligibility	RC-01. Group and eligibility structures. Released at cutover.
B	R. Lattimore	VP Claims — counterpart	RC-04. Signs the equivalence acceptance. Released at cutover + 90 days.
B	C. Adeyemi	Chief Actuary	⚠ RC-03. Named in filings; owns the measure definitions in BRD-04. Released after the first combined filing cycle.
B	M. Sheffield	General Counsel	RC-02, RC-03. Regulatory relationships and legal entity work. Released at entity consolidation.
B	J. Ferrandino	Compliance Director	RC-03. Open regulatory matters. Released at their closure.
B	N. Applewhite	Controller	RC-01, RC-04. General ledger TSA exit. Released at that exit.
B	D. Achterberg	Director, Vendor Management	RC-02. Inherited contract history. Released at vendor consolidation.
C	C. Hollifield	Director, Member Services	RC-04. Coexistence to Day 100. Released at the Day 100 decision.
C	M. Kessinger	President & CEO	⚠ RC-02. Employer, broker and regulator relationships through the transition. Released at Day 100.

4. The Asymmetry

All sixteen are at Cumberland Valley, and that is not an oversight to be corrected — it is the shape of the problem. Retention risk sits where the undocumented knowledge is, and the undocumented knowledge is at the company being absorbed. The same people whose systems the program depends on are the people whose futures the program is deciding, and they know it before anyone tells them. ⚠ They also have the strongest market position of anyone involved: a claims configuration analyst who has just been through an acquisition is exactly who a competitor wants to hire.

So the program depends for twelve to fifteen months on people it cannot offer twelve to fifteen months of certainty to. Retention payments do not resolve that. They buy time against it — which is why §7 spends that time on transfer rather than on continuity.

5. Instrument Design

Ref	Element	Provision
RI-01	Form	Cash retention agreement, separate from severance entitlement. ⚠ Accepting retention does not waive severance.
RI-02	Tranche payment, not a single cliff	⚠ Paid in installments across the retention period. A single cliff creates one date on which the program's most knowledgeable people all become free to leave.
RI-03	Release date stated	The agreement names the date, so the person is not guessing
RI-04	Transfer obligation attached	⚠ Payment is conditional on knowledge transfer being accepted, not merely delivered. See §7.
RI-05	Payable on involuntary release	⚠ If the program releases someone early, the balance is paid. Otherwise the incentive is to look busy.
RI-06	Not payable on resignation	Standard, and stated plainly at offer
RI-07	Amount set on exposure, not on level	⚠ Calibrated to what losing the role would cost the program, not to the person's grade
RI-08	Funded by Human Resources	⚠ The program identifies the roles; it does not price them. Cost sits outside the program budget.

RI-02 is the design decision most often got wrong, and its failure mode is a date rather than a person. A single lump sum at TSA exit is simpler to administer and creates exactly one problem: everyone whose knowledge the program depends on becomes free on the same morning, immediately after the milestone that proves they are no longer essential. Tranches spread that, and more usefully they give the program repeated moments where a departure is visible early rather than at the end. ⚠ RI-05 matters for the same reason in reverse — if the balance is forfeited on early release, a covered person has an incentive to appear indispensable rather than to make themselves replaceable, which is the opposite of what §7 is asking of them.

6. What Retention Does Not Buy

Not bought	Consequence and response
Engagement	⚠ A retained person can be present and disengaged. Transfer acceptance in §7 measures output, not attendance.
Discretionary effort	Retention holds a floor. Anything above it comes from how people are treated — see §8.
Loyalty after release	⚠ These are people who will leave. The plan assumes it rather than hoping otherwise.
Knowledge, by itself	Retention buys access to knowledge. Transfer extracts it. Without §7 the payment funds a longer goodbye.
Cover for the uncovered	⚠ Naming sixteen people names everyone else. See §8.

7. Knowledge Transfer Linkage

Ref	Requirement	Detail
KT-01	Every covered role has a named receiver	⚠ Named at offer, not later. A transfer obligation with no receiver is an aspiration.
KT-02	Transfer is a deliverable	Scoped, scheduled and tracked like any other deliverable in the plan
KT-03	Acceptance is by reverse shadowing	⚠ The receiver performs the work unaided while the incumbent observes. Not the other way round.
KT-04	Documentation is an output, not the test	⚠ A written runbook that nobody has executed is not a transfer
KT-05	Receiver capacity is planned	The receiving person is given time for it. ⚠ A transfer squeezed around a full workload does not happen.
KT-06	Transfer completion gates release	A covered role is not released until its transfer is accepted, or the release is escalated

KT-03 is the whole mechanism, and the conventional version of it does not work. The usual arrangement is that the incumbent demonstrates while the receiver watches and takes notes — which produces a receiver who has seen the job done and cannot do it. Reversing it is uncomfortable for both parties: the receiver is exposed doing work they are not yet good at, and the incumbent has to sit on their hands while someone slower does their job badly. ⚠ That discomfort is the test working. It is also the only version that surfaces the gap while the person who can close it is still employed, which is the entire point of having paid to keep them.

8. Communication

Moment	Owner	Approach
Offer	Line executive, individually	⚠ One to one, never in a group. The conversation is about that person's role and date.
What is said	—	The role is needed until a stated date; what happens after is not yet decided; here is what the agreement covers. ⚠ No implication of a permanent role that has not been decided.
What is not said	—	⚠ No reassurance about the wider organization. A promise made to soften a conversation is a promise the program cannot keep.
The uncovered	D. Marchbanks	⚠ Not told they are uncovered, and not told they are safe. Both would be misleading.
Ongoing	Line executive	Monthly one to one for covered roles. ⚠ A retained person with no contact concludes they are being warehoused.

The hardest sentence in this plan is the one in row two, and it is the one worth getting right: "we need this role until roughly this date, and what happens after that has not been decided." It is unsatisfying, it will not make anybody happy, and it is true. The alternative — implying continuity that has not been decided — buys three months of calm and costs every subsequent message the program sends, because the person will remember exactly what was said when the decision goes the other way. ⚠ People handle uncertainty considerably better than they handle having been managed.

9. Measurement

Metric	Target	Note
Covered roles retained to release	Above 85%	⚠ Not 100%. A plan assuming nobody leaves has no response designed.
Transfers accepted before release	100%	The measure that matters. Retention without transfer is spend without benefit.
Roles with no trained receiver	Zero by Day 100	Reported monthly to the Steering Committee
Uncovered attrition in critical functions	Monitored	⚠ A departure outside the covered list can still hurt. The list is a judgment, not a fact.
Time from resignation to receiver readiness	Tracked	The real measure of how exposed the program is

10. When Someone Leaves Anyway

Ref	Response	Detail
DR-01	Notified within 24 hours	⚠ To the Program Manager and workstream lead, not only to Human Resources
DR-02	Transfer accelerated, not abandoned	⚠ The notice period is re-planned around transfer. Ordinary handover activity is deprioritized in favor of it.
DR-03	Exposure assessed and logged	What the program now cannot do, and by when it must be able to
DR-04	Backfill decided explicitly	Replace, contract, absorb or accept — ⚠ decided, not defaulted
DR-05	Consider a consulting arrangement	⚠ A departing expert may return on contract for defined transfer work. Cheaper than the alternative and frequently refused out of pride.
DR-06	Risk register updated	The risk has materialized; it moves to issue management with an owner

The defensible summary if asked whether this program was adequately protected against key-person loss. Sixteen roles were identified by exposure rather than seniority, covered by instruments that pay in tranches and are conditional on transfer being accepted rather than delivered, each paired with a named receiver, and measured on transfers completed rather than on people retained. The plan assumes departures rather than hoping against them, and its response is written before the first one happens. ⚠ What it does not claim is that everyone stayed — that would be a claim about luck rather than about planning.

Table of Contents