When a Handback Is Actually the Right Call
A handback is justified by a structural change inside your company, not by one rough month with your provider.
Insourcing accounting is also ordinary. In Deloitte’s 2024 Global Outsourcing Survey of more than 500 executives, 70 percent said they had selectively insourced scope previously held by a third party within the last five years, while 80 percent planned to maintain or increase their outsourcing investment. The dial moves in both directions at most companies.
Four signals hold up reliably.
- A permanent controller or VP Finance is already hired. Once a full-time owner holds the close, paying an outside team to also own it duplicates the function instead of extending it.
- Volume outgrew the pricing model. Fees that scale per entity, per transaction, or per bill-pay run stop tracking value when activity climbs and the judgment content of the work stays flat.
- A system change your provider cannot run. An ERP move, a revenue recognition engine, or a multi-entity restructure can land outside their operating envelope. Architecture is forcing the change, not service quality.
- Confidentiality around a transaction. A pending sale, a financing round, or a carve-out can create a real need to tighten the circle before the numbers are public.
What isn’t a signal: a single missed close, a staffing change on the provider’s side, or a disagreement about scope. Those are service conversations, not sourcing decisions.
Two timing notes before you act on any of this. Bureau of Labor Statistics projections have the bookkeeping, accounting, and auditing clerk occupation declining about 6 percent from 2025 to 2035, and AICPA Trends data counted 55,152 accounting degree completions in 2023-24, down 6.6 percent year over year. Hiring runs slower than it used to, so sequence the hire before the notice. This page also doesn’t rerun the economics. The salary-versus-fee math and the revenue-stage thresholds live in our breakdown of outsourced accounting versus an in-house team.
What You Have to Reclaim Before You Give Notice
Ending an outsourced accounting engagement starts with the engagement letter, because you have the most bargaining power the day before notice goes out.
Open that letter and find four things first. The notice period. Whether transition assistance is included or billable. The data handover format and deadline. Who holds primary admin on each system. Then work the list below while you’re still a client in good standing.
Notice ends the easy asks.
| Artifact | Why it matters | Who usually holds it today |
|---|---|---|
| Close checklist and close calendar | The ordered task list, the day each task runs, and the dependencies. Without it the first unaided close is improvised. | The provider |
| Recurring journal entry schedules | Every standing entry, its calculation basis, its reversal convention, and who approved the methodology. | The provider |
| Reconciliation support for every balance sheet account | Bank, credit card, merchant processor, intercompany, payroll clearing, accruals, and every suspense account. The support, not the conclusion. | The provider |
| Deferred revenue and prepaid schedules | Live roll-forward files in a format that still calculates, not a flattened PDF. | The provider |
| Fixed asset and depreciation registers | Cost basis, method, useful life, accumulated depreciation, and disposals, with book and tax columns where they differ. | The provider |
| Accrual and cutoff policy | The dollar thresholds, the estimate methodology, and the rules behind the judgment calls. | Often nobody. It lives in one reviewer’s head. |
| Chart of accounts and mapping documentation | Numbering logic, department and class dimensions, and the map from ledger to board pack to tax return. | The provider |
| Vendor and customer master data | Payment terms, tax status, W-9 and 1099 flags, banking details, and approval routing. | Shared |
| Payroll mapping | Which earnings and deduction codes hit which accounts, by entity and department, plus the accrual convention. | The provider or the payroll platform |
| Admin rights on every system | Ledger, bill pay, expense, payroll, banking, state tax portals, and equity management. Confirm you hold the top-level seat. | Frequently the provider |
| The accounting file itself | A working ledger with history intact, plus flat-file exports of the trial balance, GL detail, and AR and AP aging. | You, in theory |
Admin rights carry one specific trap. In QuickBooks Online, only the current primary admin can transfer that role, and the client has to accept it through an emailed link. Intuit’s own guidance says it isn’t recommended for a firm to hold the primary admin seat. If your provider holds it, only your provider can hand it back, which is exactly why it belongs on the pre-notice list. Machine credentials deserve the same pass, because API keys, service accounts, and integration tokens routinely outlive named user logins.
What you can actually demand is narrower than most exit emails assume, and knowing the line is what makes the request land. Under the AICPA Code of Professional Conduct, the records request interpretation at ET section 1.400.200, effective July 31, 2021, a member should return client-provided records in their custody. Guidance published in the Pennsylvania CPA Journal reads the operative standard as responding expediently and generally within 45 days.
The same framework holds that working papers reflecting the member’s own testing are not client records. So don’t ask for “all workpapers.” Ask for the reconciliation support and the schedules you need in order to keep operating. That’s a different request, and a winnable one.
Tax records come back even in a fee dispute, within limits. Treasury Circular 230 at 31 CFR 10.28 requires a practitioner to promptly return any records of the client necessary for the client to comply with federal tax obligations. Where state law permits a fee-based lien, the practitioner may retain some records but still has to give you reasonable access to review and copy them.
If the engagement covered the full finance function rather than bookkeeping alone, the reclaim list runs across every layer of it. Keep what you collect, too. IRS record retention guidance sets the general period of limitations at three years, six years where unreported income exceeds 25 percent of gross income shown, and at least four years for employment tax records.
The Reverse Transition, Month by Month
Bringing accounting back in-house runs on an overlap period, a parallel close, and a named owner for every task on both sides.
Our guide to finance as a service onboarding covers the first 90 days coming in. This is the mirror image going out, and it’s the longer trip, because the receiving company has usually lost the knowledge it transferred out in the first place. Outsource Accelerator’s reverse transition plan definition puts a substantial reversal at six to 12 months end to end across five phases. The same guidance is blunt about incentives. An incumbent has little natural reason to make its own replacement smooth, so cooperation has to be a contractual duty with a price already attached, and transition assistance windows commonly run three to six months.
| Window | Phase | What has to happen |
|---|---|---|
| Month -2 to -1 | Before notice | Hire or confirm the receiving finance lead first and let that person architect the shop. Work the reclaim list. Read the exit clause. Practitioners who’ve run this recommend starting in Q2 so cutover doesn’t collide with year-end. |
| Month 1 | Discovery, 4 to 8 weeks | Document what the provider actually does, not what the scope document says. This is the phase most likely to overrun, because of workarounds nobody wrote down. |
| Months 2 to 3 | Capability build and knowledge transfer, 8 to 16 weeks and 6 to 12 weeks | Seat the team, stand up systems and access, migrate the data, and run structured transfer by process area. |
| Month 4 | Parallel close, 4 to 8 weeks | Both sides close the same period independently, then reconcile the two trial balances line by line. |
| Months 5 to 6 | Cutover and stabilization, 8 to 12 weeks | Cutover is a date. Stabilization is a phase. Keep a defined, paid channel back to the outgoing provider, with response times agreed in writing beforehand. |
Knowledge transfer has a shape worth following exactly. Documentation review first, with both sides marking gaps in the written procedures. Then shadowing, where the outgoing team runs the close while yours observes. Then reverse shadowing, where your team executes and theirs corrects. Then independent operation against a quality threshold. Most business services take six to 14 weeks per process area, and the failure mode is consistent: a team that can describe a process confidently often cannot execute it under real volume and real exceptions.
The parallel close is the month companies cut first and regret most. Run the comparison properly. Compare the trial balance account by account, then the close calendar by task completion date, then the reconciliation file list against the full account list. A parallel close that only compares net income has verified almost nothing. Every variance you surface is a documentation gap you found for free, while the people who can explain it are still on the engagement.
A reverse transition plan accounting for every recurring task, every system login, and every filing deadline is the real deliverable. In the engagements we run, that ownership map is the first artifact we build and the last one we hand over. The tasks that go missing are almost never the month-end journal entries. They’re the quarterly and annual items. Sales tax filings, 1099 preparation, state registrations, and equity bookkeeping sit on nobody’s list until they’re late.
The First Close Your Own Team Runs
The first close without your provider will run long, and that’s expected rather than a verdict on the decision.
APQC data drawn from roughly 2,300 organizations puts the median cycle time from running the trial balance to completing consolidated statements at 6.4 calendar days. Top-quartile performers finish in 4.8 days or less, and the bottom quartile needs 10 or more. A team in its first month of ownership won’t land at the median, and saying so to the board in advance is cheaper than explaining it afterward.
Three things break, and all three are ownership failures rather than technical ones.
Nobody owns the accrual cutoff. The cutoff decision is a judgment call with a dollar threshold and an estimate methodology behind it, and it usually lives in one experienced reviewer’s head. When that reviewer works for the outgoing provider, the first close either books nothing or books everything, and margin moves for reasons nobody can explain. The fix is to capture the policy during discovery as a written rule with thresholds, methodology, a reversal convention, and a named approver, then have your team apply it during the parallel close while the outgoing team can still correct it.
Reconciliations that were never actually documented. A reconciliation can be complete and still be unreproducible. The reviewer may have confirmed a balance against a source without leaving a workpaper a new person can follow. The gap surfaces when an auditor, a lender, or a diligence team asks for support on an account that has reconciled cleanly for two years. The fix is mechanical. During the parallel close, compare the reconciliation file list against the full account list, not against the accounts someone remembers. Any balance sheet account without a file is an open item, including the ones sitting at zero.
The tax calendar quietly loses its owner. This one stays invisible until a notice arrives. For returns due after December 31, 2025, the IRS failure to file penalty on Forms 1065 and 1120-S is assessed monthly for up to 12 months at $255 per partner or shareholder, and the minimum penalty on a return more than 60 days late is $525. A ten-partner partnership filing three months late is looking at $7,650 before any reasonable-cause relief, on a return that may owe no entity-level tax at all. Build the tax calendar as a separate artifact from the close calendar, with a named owner, before notice goes out. Then decide out loud whether tax moves or stays, because “we assumed it moved” is how companies collect their first notice.
One more risk, named without drama. A handback often replaces a team with a single person. If that person is the only one who knows how the close works, you’ve traded one dependency for a narrower one. Cross-training and written procedures are the answer, and the overlap period is the cheapest time to build both.
What to Keep Outsourced After the Handback
Most completed handbacks are partial, and the residual scope is predictable: tax, technical accounting, payroll, and equity support.
- Tax. Federal and state income tax preparation, multi-state nexus and registrations, and the annual provision. Episodic expertise with a concrete penalty for getting it wrong, which is why it’s the most commonly retained piece.
- Technical accounting. Revenue recognition under ASC 606, lease accounting under ASC 842, stock compensation, business combinations, and anything that needs a memo an auditor will read.
- Payroll. Multi-state payroll tax registration, filings, and reconciliations.
- Equity support. Cap table maintenance, option accounting, and the data that feeds a 409A valuation.
The logic behind that split is frequency. A controller who runs a clean monthly close isn’t necessarily the person you want tracking nexus across nine states twice a year. Low-frequency, high-consequence work is what a single in-house hire has the least time to stay current on. Clients who take the close in-house often keep the tax and technical accounting work with indinero, and that partial outcome is the most common one we see.
Decide the split during discovery, not after cutover. The full version of this model, including the three common ways companies divide the work and the places the division breaks down, lives in our guide to co-sourced accounting.
How Indinero Hands Work Back
When a client moves the close in-house, our job is to make the handover boring and complete.
Every engagement we run produces the artifacts on the reclaim list as a matter of course, not as an exit project. The close checklist and calendar. The reconciliation support for every account. The schedules, the chart of accounts mapping, the payroll mapping. A provider that can hand those over cleanly is showing you how it worked the whole time. A provider that can’t is telling you something too.
Documentation is the deliverable.
The reason the handover holds together is the shape of the engagement. Indinero runs the full finance function as one engagement, with bookkeeping, GAAP close, tax, and fractional CFO advisory from one CPA-led team, so there’s no seam between vendors where a schedule can go missing. Books kept on the accrual basis and audit-defensible from the start are also what makes a parallel close short. The support already exists and already reproduces, so your team isn’t rebuilding two years of reconciliations under a deadline.
The engagement scopes up and scopes down, year-round, as your team changes. Hire a controller and want the close inside? We hand it over and keep the work you’d rather not staff for. Continuous operations since 2009, SOC 2 compliant (2026), and a 5-star Clutch rating are why the same people are still there when you call during stabilization.
You’re not just choosing who keeps the books. You’re choosing how easy it will be to take them back. To see how the full finance function gets scoped before you decide what to keep inside, start with our outsourced accounting services.
Frequently asked questions
Most handback questions are about sequence and ownership rather than whether to do it at all. These are the ones that come up in nearly every exit conversation, answered plainly. If your situation sits outside them, the engagement letter and the reclaim list are usually where the answer starts.
What notice period is normal when ending an accounting engagement?
Your engagement letter sets the notice period for an accounting engagement, so read that clause first. No industry default overrides what you signed. Four terms matter more than the date itself. Whether transition assistance is included or billable, the data handover format and deadline, who holds primary admin on each system, and what post-notice cooperation costs. Transition assistance windows commonly run three to six months. Indinero treats the exit with the same care as the onboarding.
Which files and documentation should a provider hand over on exit?
A provider should return the close checklist and calendar, reconciliation support, journal entry schedules, deferred revenue and fixed asset registers, and system admin rights. Know where the line sits. Under the AICPA Code at ET section 1.400.200, client records come back, generally within 45 days, but the provider’s own working papers are not client records. Circular 230 requires prompt return of anything you need for federal tax compliance, even during a fee dispute. Ask for support and schedules, not every workpaper.
How long does it take to move a monthly close back in-house?
Moving a monthly close back in-house takes six to 12 months end to end under a standard reverse transition plan. The phases are discovery at 4 to 8 weeks, capability build at 8 to 16, knowledge transfer at 6 to 12, parallel running at 4 to 8, and stabilization at 8 to 12. Exits run longer than onboardings because the company has lost the knowledge it transferred out in the first place.
Who runs the close during the overlap month?
Both teams run the close during the overlap month, each closing the same period independently before reconciling the two trial balances line by line. By then your team should be in reverse shadowing, executing the work while the outgoing team corrects it. Compare account by account, then the close calendar by task completion date, then the reconciliation file list against the full account list. A parallel close that only compares net income has verified almost nothing.
What does the first in-house finance hire need to be able to do?
The first in-house finance hire needs to own the whole monthly close, including reconciliations, the accrual cutoff decision, and the tax calendar. That’s controller-level capability rather than staff-accountant capability. They should be able to write the close checklist instead of following one, document a reconciliation a new person can reproduce, set and defend cutoff thresholds, and name an owner for every quarterly and annual filing. Hire that person before notice goes out so they can architect the shop.
Can we take back part of the work and leave the rest outside?
Yes, and partial handbacks are the most common outcome, with the close moving in-house while tax, technical accounting, payroll, and equity support stay outside. That split is co-sourcing, and the logic is frequency. Multi-state nexus, ASC 606 memos, and option accounting are low-frequency, high-consequence work a single in-house hire has the least time to stay current on. Decide the division during discovery rather than after cutover. Clients who take the close inside often keep tax and technical accounting with indinero.
What usually goes wrong in the first quarter after a handback?
Three things usually break: nobody owns the accrual cutoff, reconciliations were never documented reproducibly, and the tax calendar quietly loses its owner. All three are ownership failures rather than technical ones, and the tax calendar is the expensive one. For returns due after December 31, 2025, the IRS failure to file penalty on Forms 1065 and 1120-S runs $255 per partner or shareholder per month. A ten-partner partnership three months late owes $7,650. Name the calendar’s owner before notice goes out.
