Most outsourcing regret starts the same way
A partner is underwater. Hiring fell through again. Someone on LinkedIn swears their offshore team "just works." There is a polished deck, a discounted first month, and a contract.
Then February hits.
The books come back late. The categorizations are technically in the right buckets and still wrong for the client's business. Your senior spends more time fixing than they would have spent doing the work. The assigned bookkeeper turns over. You train the replacement through your review comments. Again.
I have heard that story enough times that I treat vendor selection like fieldwork. Assumptions get tested. Marketing language gets challenged. And I care more about what happens when something breaks than what the homepage promises when everything is fine.
What went wrong for the firms that got burned
The pattern is rarely "the provider was a scam." It is more ordinary than that.
They bought staff and thought they bought delivery.
They accepted "contact us for pricing" and discovered the real number after they were emotionally committed.
They never saw a sample package that looked like something they would put in front of a client.
They confused "SOC 2 compliant" with an actual audit report.
They signed a long term before they had reviewed one real client file end to end.
None of that is mysterious. It is what happens when capacity panic outruns diligence. Accounting Today has covered how easily outsourcing fails without structure. Structure is the whole game.
Five questions every firm should ask
1. Can you review every deliverable before it goes to your client?
If the provider emails your client directly, that is not white-label. That is co-branded chaos waiting to happen.
White-label means the package comes to your firm first. Your branding. Your review. Your send.
I will be blunt. If you cannot stop a bad package from reaching the client, you do not control quality. You are renting hope.
Ask who the client talks to. Ask whose logo sits on the statements. Ask what the escalation path looks like when a number is wrong. If those answers are fuzzy, walk.
2. Who actually does the work?
"US-managed offshore team" can be a real operating model. It can also be a slogan over a revolving door.
Offshore teams with 20% to 35% annual turnover mean you are constantly retraining your provider's staff through your QA feedback. That is not a partnership. That is unpaid training services. I wrote more on that gap in outsourced bookkeeping vs. offshore staff.
I have reviewed work from offshore providers that was technically categorized correctly but missed context that any bookkeeper who understood the client's business would catch. Technically right. Practically useless.
Ask:
- Where does production sit?
- What is their retention rate on the people who will touch your files?
- Who reviews before it reaches your desk, and are they a licensed CPA?
- What happens when the assigned person quits mid-month?
The Journal of Accountancy's offshoring guide is worth reading before you sign anything that moves client data overseas.
3. What is the security posture?
SOC 2 Type II is the floor for anyone touching client financial data at volume.
Not "SOC 2 compliant." Not "SOC 2 certified." Not "we follow SOC 2 principles."
Those phrases are not a substitute for an audit report issued by an independent CPA firm. The report either exists or it does not. Ask for it under NDA. Read the exceptions. Ask what systems were in scope.
This is not paranoia. CPA firms are financial institutions under the FTC Safeguards Rule. You are responsible for service provider oversight. The Journal of Accountancy's outsourcing and professional liability piece walks through why that diligence matters when confidential client information leaves your environment.
If a provider stalls on the report, believe them. They are telling you something.
4. Is pricing transparent and flat, or does it drift?
I want a number I can put in a margin model.
Flat per-client pricing lets you forecast. Pricing that moves with transaction volume, "complexity tiers," rush fees, and mystery add-ons does not. It turns every new client into a renegotiation.
"Contact us for a quote" is not evil. It is a process cost. If every serious vendor on your list hides the number, fine, run a bake-off. If one publishes a clear rate and another makes you sit through three discovery calls to learn the price, I know which one respects partner time.
Ask what is included. Ask what triggers overages. Ask whether cleanup is separate from monthly. Get it in writing for a defined client profile, not a vague "it depends."
5. Can you test with a real client before signing a contract?
If a provider will not let you evaluate real work product, ask yourself why.
A free test client, or a paid proof-of-work on one file with an easy exit, is the only diligence that matters. Decks do not close books. Sample PDFs help. Live work on a client you know is better.
I want you to review that package the way you review your own team's work. Open items honest? Recs complete? Branding yours? Review notes usable? Partner time closer to 90 minutes than half a day?
If they need a 12-month commitment before you have seen that, they are asking you to buy blind. I do not buy blind with client data.
Red flags
Walk away, or at least slow way down, if you see these:
- No transparent pricing. Endless discovery before a number appears.
- Offshore-only with no US oversight. Especially if nobody can name who reviews final packages.
- Cannot show a sample deliverable. Or the sample is a toy file that looks nothing like a real close.
- No defined onboarding process. No intake checklist. No list of required inputs. No timeline to first clean month.
- Claims SOC 2 "compliance" instead of showing an audit report. Marketing language where evidence should be.
- Long-term contract before you have seen their work. Commitment without proof.
- They serve end clients under their own brand. Competitors are not partners.
- Your client talks to their staff. That is not white-label. That is a shared relationship you will regret at renewal time.
Green flags
These are the signals I trust more than adjectives on a homepage:
- Flat per-client pricing you can model before the sales call ends.
- US-based review by a licensed CPA with a clear handoff into your partner review.
- Real sample deliverables that look like something you would actually issue.
- Free test client or equivalent proof-of-work with no theatrics.
- SOC 2 Type II audit report available under NDA, including scope and exceptions.
- Clear onboarding checklist with defined inputs, owners, and dates.
- You remain the only client-facing contact, and nothing ships without your approval.
Notice what is missing from that list: "best-in-class," "proprietary platform," and a stock photo of a headset.
The AICPA angle people skip
Under the AICPA Code, when you use a third-party service provider, confidentiality obligations do not disappear. Interpretation guidance on disclosing information to a third-party service provider is clear that you need either client consent or a contractual confidentiality arrangement with the provider that includes reasonable procedures against unauthorized release. Many firms do both. Some state boards go further and require written client disclosure. Check your board before you assume national marketing copy is your rulebook. If the disclosure question is what is stalling you, I covered it directly in should you tell your clients you outsource bookkeeping.
Security diligence is part of ethics, not an IT side quest.
What a serious bake-off looks like
You do not need twelve vendors. You need two or three that survive the red-flag screen. For a side-by-side look at the names firms actually evaluate, see best outsourced bookkeeping services for CPA firms in 2026.
Give each the same client profile. Same entity type. Same approximate transaction volume. Same software. Same close deadline. Ask for a written all-in monthly price and a sample package for that profile.
Then run one live file. Prefer a free test client when it is offered. If the provider only offers paid proof-of-work, keep the fee small and the exit easy.
Score the output the unromantic way:
- Were reconciliations actually done, or only claimed?
- Were open items honest, or buried?
- Could a partner finish review in roughly 90 minutes?
- Did anything about the package look like another company's product?
- How many emails did it take to get a straight answer during the test?
The provider that wins the sales call and loses the bake-off is common. Believe the bake-off.
My working rule
If the provider cannot explain, in plain language:
- who does the work,
- who reviews it,
- what you pay,
- how you test it,
- how client data is protected,
then you do not have enough to sign.
Capacity pressure makes bad contracts feel urgent. I get it. I also review the cleanup after those contracts. The firms that do this well start small, keep review in-house, and refuse to confuse a lower hourly rate with a lower total cost.
Cheap production with expensive partner rework is not savings. It is deferred payroll with worse branding risk.
Ask the five questions. Demand the green flags. Treat the red flags as disqualifiers, not quirks. And do not outsource judgment. Outsource production. Keep the signature.
If the file is years behind, that is reconstruction, not a monthly close. See book reconstruction for tax professionals. For monthly white-label delivery, start at For CPA Firms. Security diligence before the first file moves is in SOC 2 and the FTC Safeguards Rule.