Rescue the book.
Start with the cause.

A rescue begins with a bounded pause, not an accusation. Preserve the current manuscript and source files. Ask everyone to stop editing separate copies. Identify the last version that the author and writer both understood to be the working draft, even if neither thought it was finished.

Then gather the proposal or agreement, the latest outline, approved sample pages, interviews, research notes and consolidated feedback. Keep these materials within the permissions and confidentiality arrangements that apply to the project. Do not upload somebody else's confidential manuscript to a new editing or AI service merely because it offers a convenient diagnosis.

Read three kinds of material side by side: what was promised, what was supplied and what was approved. A writer may have delivered the agreed outline faithfully while the author's purpose changed. An author may have provided enough material while the writer failed to use it. Both situations deserve a clear account. They need different remedies.

My broader guide to preparing before hiring a ghostwriter concerns the start of an engagement. Here, the question is narrower: what must change for an existing project to become worth continuing?

These problems can coexist. Fix their dependencies in order. A new voice pass will not rescue a chapter whose central claim has no evidence. A stronger outline will not resolve a dispute about who gets final approval.

Voice is more than favorite phrases

A convincing voice carries a person's habits of thought. Does the author reach for a story before a principle? Admit uncertainty? Interrupt a grand claim with a practical objection? Find humor in themselves or in other people? Those choices matter more than sprinkling familiar expressions into generic prose.

For a repair sample, choose material that requires judgment: an opening, a difficult explanation and an emotionally or intellectually important passage. Compare the draft with a recording or existing writing the author considers representative. Mark the places where the prose becomes more certain, sentimental, formal or self-congratulatory than the source supports.

Do not demand an exact transcript. Spoken language often needs shaping to work on the page. The useful question is whether the revision preserves the speaker's meaning and sensibility while making the writing readable. Memoir ghostwriting requires particular care here because an elegant sentence can still misrepresent how someone understands their own life.

Some projects stall because the outline is treated as a contents list rather than a set of decisions. “Leadership,” “culture” and “growth” are subjects. They do not explain why chapter three follows chapter two or what the reader should understand differently afterward.

Give each chapter one primary job. Identify the evidence, scenes or examples that allow it to do that job. Mark material as usable, incomplete, disputed or unavailable. A chapter built around an interview that cannot be obtained is a scheduling problem before it is a writing problem.

The Manuscript Revision Map can help make those structural decisions visible. Use it after agreeing on the book's purpose. If the author wants a private family history and a reviewer keeps asking for a commercial leadership book, rearranging chapters will only organize the disagreement.

Changing direction can be sensible. It should become an explicit scope decision with consequences for budget and schedule. It should not arrive disguised as a small revision to chapter eight.

An executive assistant may verify dates. A subject specialist may challenge a technical explanation. Counsel may identify legal concerns. The author may decide whether the voice feels true. Treating all four as interchangeable line editors makes the manuscript harder to finish.

Assign review roles and a person authorized to resolve conflicting editorial requests. That decision owner does not overrule specialist obligations or contractual rights. They coordinate them so the writer receives one coherent response rather than a collection of incompatible files.

Separate feedback into factual correction, structural change, voice adjustment and preference. Ask reviewers to identify the problem before prescribing the sentence. “This makes the founder sound certain before the evidence arrived” gives the writer something to solve. “Make it punchier” does not explain what is wrong.

Maintain a decision log for reopened issues. If an approved argument changes, record why and which chapters it affects. A revision allowance cannot absorb an indefinitely changing book simply because each new request arrives in a comment bubble.

The diagnostic above identifies a sensible next conversation, not a verdict on the writer. A relationship may recover if both parties can recognize the problem, supply what is missing and demonstrate improvement. Restarting becomes more reasonable when the purpose has fundamentally changed, the needed expertise is absent, the source base is unusable or trust cannot support continued collaboration.

Consider the remaining work rather than the sunk cost alone. A draft with many pages may contain little usable structure. A rough draft with sound reporting may be a strong foundation. Ask an independent editor, where appropriate, to distinguish salvageable material from material that merely looks finished.

Before changing writers, clarify rights to drafts, recordings, research and unfinished material. Payment does not by itself settle every ownership or reuse question. The Authors Guild's discussion of collaboration and ghostwriting agreements identifies ownership, control, attribution and ending the relationship as matters for written agreement. Have qualified counsel resolve project-specific disputes rather than treating an editorial assessment as legal permission.

The new brief should identify the intended reader, the book's central promise, the manuscript version, the material that can be used and the decisions already settled. It should also name the missing inputs and who will supply them.

Then specify the next deliverable, its acceptance criteria, the review process, the fee arrangement and the date on which the team will decide whether to continue. Put changes to an existing agreement in an appropriate written amendment. A rescue should reduce ambiguity, including ambiguity about what additional work costs.

Reserve author time. A writer cannot reconstruct unpublished experience from a crowded calendar and three voice notes. Equally, an author should not have to rewrite every delivered chapter to make the engagement function. The repair test should reveal whether the collaboration produces useful work at a sustainable effort level.

If AI tools are part of the workflow, address their permitted uses, confidentiality and disclosure directly. The Authors Guild's author guidance recommends written expectations when hiring collaborators. The practical issue for a rescue is knowing what material exists, how it was produced and whether it can be responsibly used.