Skip to content
Kit · Organizational design

The merger, affiliation and shared-services readiness kit

What this is: Seven rungs of combination from shared purchasing to full merger, the questions that pick the rung, the asymmetric preparation each party needs, five decision gates, the twelve things that do not transfer without somebody else's written consent, a due-diligence register, and the first hundred days.

It assumes: A board willing to discuss the question out loud. Nothing else — the first three phases cost nothing and happen before you engage counsel.

Change before use: Everything after the gate where counsel joins. Your lawyer and your auditor complete the legal and tax structure; this kit gets you to them prepared and stops where their work begins.

The problem this fixes

Two things are almost always true when a nonprofit board first says the word merger out loud. The first is that somebody has been thinking it privately for a year. The second is that the board is discussing the most binding, most expensive, least reversible option available, without having considered the six lighter arrangements that might solve the actual problem.

That happens because the vocabulary is thin. "Merger" is the only word most boards have for the whole family of ways two organizations can combine effort, so it is the word that gets used, and the conversation is immediately about identity, the executive director's job, and whose name goes on the building. Those are real questions. They are the wrong ones to start with, and starting with them is how a conversation that should have produced a shared back office instead produces eighteen months of anxiety and nothing else.

This kit does three things. It gives you seven rungs instead of one word, so the board can choose the lightest arrangement that solves the problem. It tells you what to do before you spend a dollar on counsel, which is more than most boards expect. And it names, specifically, the things that do not transfer when organizations combine — the licenses, contracts, accreditations, and coverages that quietly need somebody else's written permission, and that are the actual reason combinations in this sector fail late and expensively.

What this is not — read this before the board sees any of it

This is not legal, tax, or accounting advice, and nothing in it is a form of agreement. It contains no draft terms, no model clauses, and no template contract, on purpose. Combining two charitable organizations is a transaction with statutory, regulatory, tax, employment, and fiduciary dimensions that vary by state and by what each organization holds. Your own counsel and your own auditor complete the legal and tax structure. This kit exists to get you to them prepared, and it stops at the point where their work begins.

It is not a valuation method. Nobody owns a charity, so there is no purchase price, no equity, and no consideration in the ordinary sense. The section on value explains what is being weighed instead, and it is not a number.

It is not an insurance opinion. Coverage questions raised here — particularly what happens to claims-made policies — are flagged so that you raise them early with the licensed broker you engage. They are not answered here and cannot be.

It is not a recommendation that you combine. A well-run process that ends in "no" is a success, and the last section is about how to end one well. The pressure in these conversations runs almost entirely one way, toward proceeding, and a kit that only described how to proceed would add to it.

The ladder — seven rungs, not one word

These are ordered from least binding to most. Each rung solves a narrower problem than the one below it and costs less to enter and to leave. The discipline this section is asking for is simple: name the problem first, then take the lowest rung that solves it. A board that starts at rung seven and works backwards almost never gets back down.

RungWhat it isWhat it solvesWhat it costs you
1 · Coordinated practiceTwo organizations agree to align something — referral protocols, a training calendar, a shared waiting list, common outcome definitions. No money changes hands and no document binds anyone.Duplication and friction at the edges where you already touch each other.Almost nothing, which is also its limit. It survives only as long as both executives want it to.
2 · Shared purchasingBuying together — insurance, benefits, training, software licenses, audit services — to reach a scale neither has alone.Unit cost, and sometimes access to something neither could buy at all.A written arrangement about who signs, who pays, and what happens when one party leaves mid-term.
3 · Shared staff or shared functionOne person or one function serves both — a shared finance director, a shared compliance officer, a shared IT function. One organization employs, the other reimburses.A capability neither can afford at full time, and the key-person exposure of having none.A real agreement covering employment, supervision, cost allocation, confidentiality, and what happens to the person if the arrangement ends. This is the rung most often done on a handshake and most often regretted.
4 · Shared services organizationA back office serving both — finance, HR, IT, compliance — either housed in one party or in a new entity the parties create.The whole administrative cost structure, permanently, and usually the quality of it too.Substantial. Governance of the shared entity, service levels, a cost-allocation methodology both boards accept, and an exit that does not leave either party without a back office.
5 · Joint program or joint ventureA specific program run together, often to pursue a contract neither could hold alone.Access to funding and scale for one line of work, without touching anything else.Clear allocation of risk, liability, and licensure — and clarity about which organization is the licensed or contracted party, because usually only one can be.
6 · Affiliation or parent–subsidiaryBoth organizations continue to exist, with one holding defined reserved powers over the other — commonly budget approval, executive appointment, and reserved decisions named in the governing documents.Strategic alignment and, often, a distressed organization's survival, while keeping licenses, contracts, and identity intact under the existing entity.Genuine loss of autonomy for the subordinate party, and a governance structure that has to be lived rather than filed. Often the right answer where licenses make a merger dangerous.
7 · Merger or consolidationOne entity survives and the other ceases to exist, or both cease and a new one is formed. One board, one budget, one set of books.Everything the lower rungs solve, permanently, plus the duplicated governance and administrative overhead that no lower rung removes.The most, and the least reversible. It is also the rung where the things that do not transfer are most likely to be discovered too late.

Two observations that hold across this ladder and are worth saying to a board explicitly. The first: rungs three and four solve the problem most boards actually have. The problem is usually administrative cost, thin capability, and key-person exposure — none of which requires either organization to stop existing. The second: rung six exists largely because of licensing, and in a sector where a license is the right to operate at all, an arrangement that leaves both licenses undisturbed is not a compromise. It is frequently the more competent answer.

The three questions that pick the rung

  1. What is the problem, stated without the word merger in it? Write it in one sentence. If the sentence is "we cannot afford a finance director", the answer is rung three. If it is "our administrative cost is unsustainable at our size", it is rung four. If it is "we will not survive the year", it is six or seven, and the timeline section below matters more than anything else in this kit.
  2. What must not change? List it before you talk to anyone. Licenses, a specific contract, a program's identity in a community, a donor relationship, a building. This list is what rules rungs in and out, and the items on it are usually the ones nobody thought to say aloud.
  3. How long do we have? A combination pursued from strength takes a year to eighteen months and can be stopped at any gate. One pursued from a cash cliff takes whatever time is left and cannot be stopped, which removes your ability to walk away — and the loss of that ability is itself the largest cost in the whole exercise.

The asymmetry

Almost every discussion of nonprofit combination is written as though the two parties are doing the same thing. They are not. The organization in the stronger position and the organization in the weaker one face different risks, need different preparation, and are usually failed in different ways. Treating the process as symmetrical is the most common structural mistake, and it is invisible from inside either party.

The stronger partyThe weaker party
The real riskAcquiring an obligation you did not price — a lease with years to run, a contested claim, a repayment demand, a licensing history, a deferred maintenance bill on a building.Losing the thing that made the organization worth preserving — a program identity, a community relationship, staff who stay for the culture — while the promises that protected it were verbal.
What preparation meansDiligence discipline. A written request register, a named reviewer per domain, and the willingness to stop at a gate when a finding is material rather than absorbing it because momentum has built.Getting your own house in order before anybody looks. An organization that cannot produce its own contracts, minutes, and financials is negotiating from a weaker position than its actual condition warrants.
The conversation to have earlyWhat are we not willing to take on, and what would we do if we found it at gate four rather than gate two?What must survive this, and is it written into the structure or is it a promise from a person who may not be here in three years?
The failure modeAbsorbing an organization and its problems, and being materially worse off two years later while nobody will say so.Agreeing to terms because the alternative is closing, and discovering that "nothing will change" was sincerely meant and not enforceable.
What good looks likeA decision made at a gate, on written findings, that the board could defend to a regulator or a donor a year later.Protections in the governing documents rather than the transition plan — because a transition plan expires and a reserved power does not.
For the weaker party specifically

The single most useful thing a board in the weaker position can do is start early enough to still have a choice. Every protection available to you — a named program, a reserved board seat, a commitment about staff, a restriction honored — is negotiable while you can still walk away and unavailable once you cannot. This is why the timeline question is the third of the three questions and not an afterthought.

The second most useful thing is to be able to produce your own records. Not because anyone is judging your filing, but because a diligence process that turns up gaps reads as risk, and risk is priced into terms. An organization that hands over a complete file is treated differently from one that assembles it under pressure over six weeks.

Phase 0 — everything you can do before spending a dollar on counsel

Boards routinely engage counsel at the point where they should be doing the following, which costs nothing and takes about six weeks. Doing it first means the legal work when it starts is narrower, faster, and materially cheaper, because your lawyer is answering questions rather than gathering facts.

  1. Write the problem statement without the word merger in it. One sentence, agreed by the board, not the executive alone.
  2. Write the must-not-change list. Both parties do this separately and do not show each other until both are written. Comparing two lists written independently is the single most informative hour in the whole process.
  3. Assemble your own file. The due-diligence register later in this kit is the list. Assemble yours before anyone requests it, whichever party you are. The gaps you find are your findings, discovered privately and cheaply.
  4. Build the does-not-transfer register. Every license, contract, accreditation, lease, grant, and policy that would need somebody else's written consent. This is the section that most changes which rung is viable, and it is entirely researchable from documents you already hold.
  5. Have the conversation about people, in the abstract, before it is about names. Not "who will be executive director" but "how would that be decided, and by whom, and when". Deciding the method before the candidates are visible is the only chance you get to decide it cleanly.
  6. Agree the confidentiality position and who knows. Including what staff will be told and when — the answer "nothing until it is decided" is a decision with consequences, not a neutral default.
  7. Agree what would make each board stop. In writing, before the process starts, while stopping is still hypothetical and therefore easy to say out loud.

Only then engage counsel — and engage them on a defined question, which is what the first gate produces.

The five gates

A combination should be stoppable at five specific points, each with a written decision by both boards. Gates matter because momentum in these processes is one-directional and largely invisible: nobody ever announces that the process has become unstoppable, it simply becomes true, usually somewhere between gates three and four. Naming the gates in advance is how a board keeps the ability to say no.

GateThe questionWhat must exist to pass itCost of stopping here
1 · ExploreIs there a problem worth solving together?Both problem statements, both must-not-change lists, and an agreed rung to explore. No counsel yet.Almost none. Two boards had a conversation.
2 · StructureIs the rung we picked actually available?The does-not-transfer register, complete, with the consents each item would require identified. Counsel engaged on a defined question. A preliminary answer on whether licenses and major contracts survive the structure.Low. Some legal fees, and a much better understanding of your own obligations, which you keep either way.
3 · DiligenceDoes what we found change the answer?A completed diligence register with findings and risk ratings, from both directions — including the stronger party disclosing to the weaker one, which is routinely skipped and should not be.Moderate, and this is the last gate where stopping is straightforward. If the board is going to lose its nerve about stopping, it happens after this point.
4 · TermsDo the protections that matter live in the documents?The governance structure, the reserved powers, the board composition, the executive decision method, the name, and every commitment either party is relying on — in the governing documents, not in a transition plan.High, and staff usually know by now. Stopping here is painful and is sometimes still correct.
5 · Close and integrateCan we operate on day one?Every consent obtained in writing. Payroll, on-call, licensing notifications, insurance, and records access all tested against the closing date, not assumed.Stopping is no longer really available. This gate is about not breaking service, and it is an operational gate, not a strategic one.

What does not transfer

This is the section that earns the kit. Boards reason about combination as though the surviving organization inherits everything, because in commercial transactions that is broadly how it works. In this sector it is frequently not, and each of the following is a separate item requiring a separate consent from a separate third party on a separate timeline. Any one of them can determine which rung is possible, and several of them take longer to obtain than the entire rest of the process.

WhatWhy it does not simply carry overWhat to do about it, early
Licenses and permits to operateA license is typically issued to a specific legal entity. Where an entity ceases to exist, its license may cease with it — and a successor may face a new application rather than a transfer, with the timeline and the site, staffing, and approval requirements of a new application.Establish, in writing from the licensing authority, what happens to each license under each structure being considered. Do this at gate two. It is the single most common reason a merger becomes an affiliation.
AccreditationAccreditation attaches to an organization and a scope of service. A structural change is usually a reportable event and may trigger review, a site visit, or re-application.Read the accreditor's change-of-status provisions before choosing the rung, and notify on their timeline rather than yours.
Government contractsMost carry assignment or change-of-control provisions requiring prior written consent, and consent is discretionary. Some carry re-procurement triggers.Read every contract for its assignment clause and build the consent list. Approach contracting officers early, and expect the conversation to be slower than commercial equivalents.
Restricted funds and donor intentA restriction follows the gift, not the entity. Restrictions do not dissolve because the organization did, and some instruments name the organization specifically.Inventory every restriction with its instrument. Where an instrument names the entity, that is a question for counsel at gate two, not a detail for closing.
Grant awardsMany require prior approval of a change in legal status, and some treat it as a termination event. Federal pass-through awards carry their own requirements.Per-award review. The Open Shelf grant and contract award inventory is the instrument; if you do not have one, this is the reason to build it.
Real propertyProperty acquired with public or philanthropic funds frequently carries use restrictions, reversionary interests, or repayment obligations triggered by a change in ownership or use.Title review on every property, looking specifically for funder-imposed restrictions rather than only encumbrances. This is a place where an unpleasant surprise is expensive and late.
LeasesMost contain assignment and change-of-control clauses, and landlord consent may be conditioned.List every lease with its clause and its remaining term.
Insurance — particularly claims-made policiesA claims-made policy covers claims *made* while it is in force. When an entity ceases to exist and its policy ends, claims arising later from earlier conduct may be uncovered unless extended reporting coverage is arranged.Raise this at gate two with the licensed broker you engage. Do not leave it to closing: it is a coverage question with a cost, and both are better known early. Nothing in this kit answers it.
Retirement plansPlan mergers, terminations, and transfers carry their own regulatory requirements and timelines, and defined-benefit or multiemployer arrangements can carry withdrawal exposure.Identify the plan type at gate two and get it in front of counsel and the plan's own advisers immediately. Timelines here are long.
Employee tenure, accrued leave, and termsWhether service is recognized, and whether accrued leave transfers or is paid out, is a decision with a cash cost — not an automatic consequence.Quantify accrued leave on both sides at gate three. It is real money and it is routinely omitted from the financial picture.
Client and personnel recordsRecords carry retention obligations, access restrictions, and sometimes consent requirements that follow the record rather than the organization. Custody must be resolved, including for programs that do not continue.Decide custody, access, and retention for every record class before closing — including records of discontinued programs, which have obligations but no owner once the program ends.
The name, and goodwillA name that carries community trust does not transfer by operation of structure alone; using it, retiring it, or preserving it as a program identity is a decision, and one that people care about more than boards expect.Decide it at gate four, in the documents. "We will keep the name" is a promise that outlives the person who made it only if it is written down.
The rule this section produces

Every item above is a separate consent from a separate party on a separate timeline, and the process moves at the speed of the slowest one. Build the register early, put a date and a named owner against each consent, and treat the longest timeline as the real schedule. A board that sets a closing date before this register exists has set a date it does not control.

Due diligence — what to request, and why each item matters

The register below is organized by domain. Two disciplines make it work, and neither is about the list. First, diligence runs in both directions. The weaker party requesting the stronger party's financials, licensing history, and litigation position is not impertinent; it is the only way its board can discharge its own duty. A stronger party that declines to reciprocate has told you something. Second, every request has a named reviewer who is competent to read it. A file that is received but not read by someone who understands it has produced paperwork, not diligence.

DomainWhat to requestWhat you are actually looking for
GovernanceArticles, bylaws with all amendments, minutes for three years including committee minutes, conflict-of-interest policy and current declarations, board roster with terms.Whether decisions were made properly and can be evidenced — and whether the bylaws contain anything that constrains the structure you are contemplating.
FinanceAudited statements and management letters for three years, current-year interim statements, budget, cash flow, the full list of restricted funds with their instruments, debt and guarantees, related-party arrangements.The management letters more than the statements. And restricted funds, which are the most commonly understated item in a nonprofit balance sheet discussion.
Contracts and grantsEvery funding agreement, with assignment and change-of-control clauses identified; the reporting calendar; open findings and corrective actions.Which agreements survive the structure, which need consent, and which are already carrying an unresolved finding.
Licensing and complianceCurrent licenses, three years of inspection reports and corrective action plans, complaint and incident history, any pending regulatory matter.Pattern rather than incident. Three unrelated findings are noise; the same finding three times is a system that does not work.
WorkforceRoster with hire dates, compensation and classification, accrued leave balances, employment agreements and severance commitments, benefit plans, open claims and grievances, personnel policies.Accrued leave and severance commitments — real, quantifiable, and routinely absent from the financial picture until late.
Property and assetsDeeds and title reports, leases, funder restrictions on property, equipment schedules, deferred maintenance assessments.Reversionary interests and use restrictions, and a maintenance backlog that is a real liability nobody has written down.
Insurance and claimsAll policies with limits, retentions, and whether each is occurrence or claims-made; five years of loss runs; open and closed claims; any coverage declination.The occurrence-versus-claims-made distinction, which determines whether a coverage gap opens when an entity ends. Take it to the broker, not to this kit.
Technology and dataSystems and vendor inventory, data holdings by class, contracts with data provisions, security incidents, retention schedule.What data is held, under what obligations, and whether it can lawfully move. The Open Shelf cybersecurity self-assessment covers this domain in depth.
Programs and outcomesProgram descriptions with capacity and utilization, outcome data and its definitions, waiting lists, referral sources.Whether the programs do what both parties believe they do, measured the same way. Two organizations rarely define the same indicator identically.
Litigation and contingenciesPending and threatened litigation, regulatory investigations, unasserted claims counsel is aware of, indemnities given.Exactly what it says. Ask the question in writing and take the answer in writing.

Every item in the register carries four fields that matter more than the item: requested, received, reviewed by whom, and the finding with a risk rating. A register showing forty items received and nine reviewed is not eighty percent complete. It is twenty-two percent complete, and the workbook computes it that way deliberately.

Cultural diligence, done concretely

Cultural fit is the thing everyone says matters and almost nobody assesses, because the usual method — asking whether the cultures are compatible — produces the answer yes from people who want the answer to be yes. Ask about mechanisms instead. These are observable, and the answers differ sharply between organizations that describe their cultures identically.

  • How is a serious mistake handled? Ask each organization to describe the last one, without names. The difference between "we found the system that allowed it" and "we addressed it with the individual" is the whole of the answer.
  • Who can spend fifty dollars without asking? A concrete question about where authority actually sits, which no mission statement will tell you.
  • How long from a decision being made to front-line staff knowing about it, and by what route? Two organizations with the same values can differ by weeks, and after combination the slower one's staff will experience the faster one as chaotic.
  • What happens when a funder asks for something the organization thinks is wrong? This predicts a great deal about how the combined organization will behave under pressure.
  • What did the last three people who left say on their way out? And whether anyone asked.
  • Where do the two organizations disagree about the people they serve? There is always a difference and it is always substantive. Finding it before combination is far better than discovering it in a supervision meeting afterwards.

Value, where nobody owns anything

There is no purchase price, because there is no owner. What each board is weighing is not consideration but a bundle of contributions and obligations, and the honest way to handle it is to write both columns down for both parties rather than to reach for a number that does not exist.

What each party bringsWhat each party brings with it
Unrestricted net assets and liquidityDebt, guarantees, and lease commitments
Contracts, grants, and funder relationshipsOpen findings, corrective actions, and repayment exposure
Licenses, accreditation, and the right to operateLicensing history and any pending regulatory matter
Programs, capacity, and demonstrated outcomesPrograms that lose money and would need to be carried or closed
Staff, and the specific expertise that is hard to replaceAccrued leave, severance commitments, and open claims
Property, and equipment that worksDeferred maintenance, use restrictions, and reversionary interests
Name, reputation, and community standingAnything that would be uncomfortable on a front page
Restricted funds, where the restriction can be honoredRestricted funds where it cannot, or where the instrument names the entity

Write both columns for both organizations, including your own. A board that has honestly listed its own right-hand column negotiates differently, and better, than one that has only listed the other party's.

The question is not what the other organization is worth. It is what each of us brings, what each of us brings with us, and whether the combination serves the people we exist for better than two organizations do.

The decisions that actually end these processes

Combinations rarely fail on diligence findings. They fail on three decisions that are personal, are usually deferred, and become harder every month they are deferred.

  1. Who leads. Decide the *method* at gate one, before the candidates are visible — a search, an appointment by the combined board, a defined transition. Deferring this until gate four means deciding it when both executives know what the answer costs them personally, and it is the most common point of collapse. Whatever is decided, the person not selected deserves a real answer about their position, in writing, before the vote and not after.
  2. Who governs. Board composition, and — more importantly — the reserved powers. A weaker party given four seats on a nineteen-person board has representation and no protection. If something must be protected, protect it with a reserved power over that specific decision, not with seats.
  3. Whose name. Sounds like branding and is not. A name carries community trust, donor relationships, and staff identity, and dropping one is a real loss to real people that should be acknowledged rather than managed. Decide it in the documents at gate four, and decide what is preserved: sometimes the right answer is that the name continues as a program identity.

Day one, day thirty, day one hundred

The integration section of most combination plans is aspirational. This one is not: the day-one list is only the things that break loudly if they are wrong, and each of them can be tested against the closing date in advance rather than discovered on the morning.

WhenWhat must be trueHow you know before the day
Day oneEveryone gets paid. The on-call phone rings somewhere and a person answers. Staff can reach the records they need to do today's work. Every license remains valid and every required notification has been made. Insurance is in force with no gap. Clients and families have been told, by someone they know.Each item tested in advance: a parallel payroll run, a test call to the on-call number, a named person confirming record access, written confirmation of each notification, and the broker's written confirmation of coverage from the closing date.
Day thirtyEvery consent obtained is filed and every one outstanding has an owner and a date. One finance function, or a written plan with a date. Nobody is unsure who their supervisor is. The reporting calendar is merged and nothing has been missed.A single consent register, reviewed weekly, and a merged obligations calendar built before closing rather than after.
Day one hundredOne set of policies, or a dated plan to get there. One set of outcome definitions. The commitments made in the documents are being kept and somebody is checking. Staff who were going to leave have mostly left, and that is known rather than discovered.A hundred-day review presented to the combined board in writing, naming what has not been done as well as what has.

At day three hundred and sixty-five, ask the question the whole exercise was for: is the combined organization serving the people it exists for better than two organizations did, and how would we know? Answer it in writing to the board. If the honest answer is not yet, that is useful and survivable. If the honest answer is that nobody can tell, the combination was never measured against anything, and that is worth knowing before the next one.

Walking away well

A process that ends in no is a success if it ends cleanly, and the pressure in these conversations runs almost entirely toward proceeding — sunk cost, staff who already know, boards that do not want to have wasted a year, and the plain awkwardness of telling a partner organization no. That pressure is why the gates and the stopping conditions are written down at the start, when saying them is easy.

  1. Say it at a gate, on written findings. A decision that points to a finding is a decision. One that points to a feeling will be relitigated for a year.
  2. Tell the other board directly and first, before staff, funders, or anyone else hears it. In person if that is possible.
  3. Say what would have to change, if anything would. Sometimes the answer is a lower rung, and rungs one to four often survive a no at rung seven — they may even be the better outcome the process was always going to find.
  4. Handle the information. Return or destroy what was shared, in writing, on the terms agreed at the start. This is the point at which the confidentiality agreement earns its existence.
  5. Tell staff something true and soon. They know. What they do not know is whether it means their job is safe, and silence answers that question badly.
  6. Write down what was learned. Both organizations now hold a complete picture of their own obligations, restrictions, consents, and gaps that they did not have a year ago. That is a real asset, and it is the reason a well-run process that ends in no is still worth having run.

Companion resources on the Open Shelf

  • Grant and contract award inventory — the instrument behind the consent register. Build it first if you cannot currently list your awards and their assignment provisions.
  • Systems and vendor inventory — the technology and data domain of diligence, and the contracts that carry data provisions.
  • The cybersecurity and data protection self-assessment — records custody, retention, and what may lawfully move.
  • The financial sustainability model — the four numbers each board should be able to state about itself before it is asked, and the concentration figure that often starts these conversations.
  • The enterprise risk register and board risk appetite statement — where a board records what it is and is not willing to take on, which is the discipline gate three depends on.
  • The executive transition and key-person risk plan — the leadership decision, and the interim arrangements a combination frequently requires.
  • The accountability and ownership map — day thirty, when nobody should be unsure who their supervisor is.

All of them are free, require no registration, and are yours to adapt. None of them, individually or together, is legal, tax, or accounting advice, and none replaces your own counsel and auditor — who complete the structure this kit stops short of.

Free to adapt and use. No registration, no fee, no attribution required. This is a working template, not legal, tax, or accounting advice — the requirements that apply to your organization are confirmed by your own counsel, auditor, and licensing authority.

← Back to the Open Shelf