Sample shortlist · candidates written by us, real screener output
Bookkeeper
This is what every search delivers: finalists screened against the brief, each with a score, screening notes, structured answers, and a graded work sample. These sample candidates were written by us and run through the real screener — the scores and notes are its unedited output, not a mock-up.
The brief they were screened against
Role: Bookkeeper · Band: $900–1,500/mo
Must-haves: monthly close on QuickBooks Online or Xero, bank and payment-processor reconciliation, e-commerce revenue (Shopify/Stripe), a real method for hunting discrepancies
Nice to have: AP/AR chasing, sales-tax filings, multi-entity
Scored by the same rubric the production screener runs on real applicants (bookkeeper, claude-sonnet-4-6), on 2026-09-20. Nothing below was edited afterwards — including where it went against a candidate. The rubric and weights are published.
Rhea M.
SCREEN 88/100Screening notes
Strongest concrete evidence: the Stripe gross-deposit misclassification catch ($38K over 11 months) is a real, named, quantified finding that demonstrates exactly the discrepancy-hunting instinct this role demands; the scenario answer is textbook-quality — divisibility-by-9 test, half-of-a-number check, and the processor-fee reconciliation path are all legitimate techniques presented in correct priority order, not recited from a list. The work sample shows genuine architectural judgment: merchant fees in COGS not expense, per-channel ad sub-accounts, A2X as the bridge layer, and the closing sequence is operationally sequenced rather than generic — these are decisions only someone who has actually cleaned up e-commerce books would make unprompted. Biggest concern: all claims are self-reported with no verifiable artifact (no anonymized reconciliation output, no actual QBO screenshot, no client reference structure), and the work sample, while excellent, could theoretically be composed rather than demonstrated; at interview, a live reconciliation exercise on a seeded file would be the only real verification. Verdict: advances to final interview; score sits at the high end of the top-10% band because the scenario and work-sample answers contain specific professional heuristics that generic applicants do not produce, but stops short of 90 because nothing is independently verified.
Experience
6 years bookkeeping for US clients, the last 4 at a Manila outsourcing firm where I carried 7 e-commerce accounts doing $200K–1.2M a year each. Monthly close on QuickBooks Online for all seven, closing by the 8th business day. Reconciled Shopify, Stripe, PayPal and Amazon settlements against the bank every month. Caught and cleaned an 11-month misclassification of Stripe fees at one client that had overstated revenue by about $38K cumulatively — the prior bookkeeper had been booking deposits gross.
Scenario answer
A $2,340 gap is a number I would want to recognise before I go hunting, because most of them are one of four things and I can usually name it in ten minutes. First: is it exactly a transaction? I sort both sides by amount and look for 2,340.00 sitting alone on one side. If it is there, it is a missing entry or a duplicate and I am done. Fastest check, so it goes first. Second: is it divisible by 9? 2,340 divided by 9 is 260, so yes — that is the classic transposition signature. I would look for a figure entered as 2,340 that should be 3,240, or 340 that should be 304-ish. Trial-balance-era trick and it still works because humans still type. Third: is it half of something? 2,340 doubled is 4,680. If a 4,680 entry went in on the wrong side of the ledger, the books move by exactly twice the amount. I search for 4,680. Fourth, and for e-commerce this is where it usually ends: processor fees. Stripe pays out net, Shopify Payments pays out net, PayPal nets per-transaction. If a deposit was recorded at the gross sale amount and the fee was never booked, the difference IS the fee. On a $200K/month store, $2,340 is right in the range of one month of Stripe fees at 2.9% on roughly $80K. I would pull the Stripe payout report for the month, total the fees column, and compare it to whatever hit the fee expense account. If all four miss, I stop guessing and start bisecting: reconcile the first half of the month alone, then the second, and keep halving until the gap lives in one week, then one day. Slower, but it always terminates. What I would not do is post a plug to Ask My Accountant and move on. That is how a $2,340 gap becomes a $2,340 gap every month for a year.
Work sample
CHART OF ACCOUNTS + CLOSE CHECKLIST — $50K/month e-commerce TOOLS QuickBooks Online Plus (needs class tracking and inventory). A2X to bridge Shopify and Stripe into QBO — this is the piece most $50K/month stores skip and it is the single biggest source of the reconciliation pain later. Dext for receipt capture. Bill.com only once AP is over about 15 bills a month; below that it is overhead. CHART OF ACCOUNTS — the parts that matter Income 4000 Product sales 4100 Shipping income (separate — it is revenue, and you need it separate to see true shipping margin) 4200 Discounts and refunds (contra) Cost of goods sold 5000 Product cost 5100 Inbound freight and duty 5200 Fulfilment and pick-pack 5300 Merchant fees (Stripe, Shopify Payments, PayPal, Amazon) Expenses 6000 Advertising — one sub-account per channel (Meta, Google, TikTok). Merged, you cannot answer the only question the owner will ask. 6100 Software 6200 Contractors 6300 Payroll and payroll taxes 6400 Owner draws (equity, not expense — this is the mistake I see most) The two decisions that cause the most cleanup later: keep merchant fees in COGS rather than expense so gross margin is real, and never let ad spend sit in one account. MONTHLY CLOSE CHECKLIST Day 1–2 1. Confirm all bank and credit-card feeds pulled through the last day of the month. A feed that stopped on the 27th is the most common cause of a mystery variance. 2. Import A2X settlement summaries for Shopify and Stripe. Confirm each settlement maps to a bank deposit — amount and date. 3. Enter or import any bills not in Bill.com. Day 3–4 4. Reconcile every bank and card account. Difference must be $0.00, not close. 5. Reconcile the processors: gross sales less fees less refunds must equal net deposits for the month. This is the step that catches fee misbooking. 6. Inventory: compare units sold in Shopify to COGS posted. If the store uses average cost, recompute; if it uses landed cost, confirm freight was allocated. Day 5–6 7. Accruals — any invoice for the month received after month end. 8. Review P&L against prior month line by line. Anything moving more than 20% gets explained in writing before close, not after a question. 9. Review balance sheet: Undeposited Funds should be zero or explainable, Ask My Accountant should be empty, owner draws should not be in expenses. Day 7–8 10. Lock the period in QBO and send the owner three things: P&L, balance sheet, and a short note naming what moved and why. THE ONE CHECK THAT CATCHES THE MOST Step 5. On an e-commerce book, if net deposits tie to gross sales minus fees minus refunds, almost nothing else can be badly wrong. If it does not tie, stop and find it before doing anything else — every downstream number is built on it.
Daniela O.
SCREEN 84/100Screening notes
Strongest evidence: three-entity Xero close with intercompany eliminations, named tools (A2X, Hubdoc, Bill.com, Fathom), a real AR metric (47→31 days DSO with a specific mechanism), and a discrepancy-hunting sequence that goes divisibility-by-9 → half/double → week-by-week narrowing — that is a practitioner's method, not a textbook list. The e-commerce work sample is structurally sound: sales-tax-payable-as-liability note shows she has cleaned up that specific error twice, and naming the gross-to-net processor rec as the load-bearing step is the right answer. Biggest concern: no direct Shopify/A2X live experience is confirmed beyond naming A2X in the sample (she says 'I would use A2X' rather than 'I do'), and the salary ask of $1,450 sits at the top of band, which is fine only if the client values the multi-entity depth. Verdict: advances to interview without hesitation; not 85+ because the e-commerce tool claims are partially conditional and nothing in the answers genuinely reframed the problem.
Experience
7 years in accounting, the last 3 as the sole bookkeeper for a US agency group with three entities (an LLC, an S-corp and a small holding company) on Xero, plus intercompany eliminations at year end. Monthly close by the 10th across all three. Handled AP for ~60 vendor bills a month and AR chasing on ~$180K of open invoices, which I brought from 47 days average collection to 31 over about eight months by doing the boring thing: an invoice goes out the day the work is signed off, and a reminder goes at day 15, not day 30.
Scenario answer
My order is cheapest test first, and the cheapest test is always the one that needs no thinking. 1. Confirm the statement period actually matches the reconciliation period. A gap that appears out of nowhere at month end is very often a statement that runs to the 28th against books that run to the 30th. Thirty seconds to check, and it is the answer more often than it should be. 2. Search for the exact amount on both sides. 2,340.00. If it exists on one side only, it is a missing or duplicated entry. 3. Divisibility by 9 — 2,340 is divisible by 9, which points at a transposition. I would scan for two digits swapped in the right magnitude range. 4. Look for 1,170 (half) and 4,680 (double). Half means something was entered once instead of twice or the split was wrong; double means an entry hit the wrong side. 5. If none of that lands, I narrow by date. Reconcile week by week until the variance appears in one week, then day by day. I would rather spend forty minutes on this than post a plug. 6. For this client shape I would also check uncleared items specifically — a cheque written in a prior month that cleared this month, or a deposit in transit at the cut-off. These are not errors, they are timing, and if that is the answer the fix is a reconciling item, not a correction. When I find it, I document what it was and why it happened in the close notes. If the same cause shows up twice, it is a process problem rather than a mistake, and I change the process — usually by moving the step earlier in the close.
Work sample
CHART OF ACCOUNTS + CLOSE CHECKLIST — $50K/month e-commerce TOOLS Xero with A2X, Hubdoc for documents, Bill.com for AP once volume justifies it. I would use Xero over QuickBooks here only if the client already has an accountant on Xero; in the US the accountant network is mostly QuickBooks and that matters more than the software. CHART OF ACCOUNTS Revenue: product sales, shipping income, discounts and refunds as a contra account. COGS: product cost, inbound freight and duties, fulfilment, payment processing fees. Operating expenses: advertising split by channel, software, contractors, payroll and payroll taxes, professional fees, bank charges. Balance sheet items that need care: inventory, undeposited funds, sales tax payable, owner's equity and draws. Sales tax payable deserves its own note. On an e-commerce book it is not an expense and it is not revenue — it is money held for a state. Booking it into income is a mistake I have had to clean up twice and it makes every revenue number wrong until you do. MONTHLY CLOSE CHECKLIST 1. Verify all feeds are current through month end. 2. Import processor settlements and match each to a bank deposit. 3. Enter outstanding bills and expense claims. 4. Reconcile all bank and card accounts to zero difference. 5. Reconcile processor gross-to-net: sales less fees less refunds equals deposits. 6. Post inventory and COGS for units sold. 7. Review sales tax collected against sales tax payable and confirm the filing is scheduled. 8. Accruals and prepayments. 9. Review P&L and balance sheet for anything unexpected; investigate before closing, not after. 10. Lock the period and issue the reporting pack with a written commentary. THE CHECK THAT CATCHES THE MOST Step 5, the gross-to-net processor reconciliation. If that ties, the revenue line is right, and the revenue line is what everything else hangs off.
Jonalyn C.
SCREEN 66/100Screening notes
Strongest concrete evidence: candidate names real tools (QuickBooks Online, Dext, Excel), demonstrates e-commerce awareness by specifically calling out Shopify and Stripe payout matching in the close checklist, and the chart of accounts correctly separates merchant fees into COGS — a non-obvious choice that shows some real exposure. Biggest concern: the experience is consistently described in supporting/subordinate terms ('for the senior to review and sign off,' 'give the reports to the client or senior accountant') with no owned close, no client names, no transaction volumes, and no numbers anywhere — this is an accounting assistant who has watched a close, not run one; the scenario answer is structurally sound but entirely procedural and generic, containing zero specific tools, thresholds, or real examples of a discrepancy she actually found. The work sample is competent and better than modal but reads as a clean constructed template rather than evidence of a live process — missing items like accruals, prepaid amortization, payroll reconciliation, and AR/AP aging review that a practitioner who has closed independently would instinctively include. Verdict: advances cautiously to a screening interview focused specifically on whether she can own a close solo; at $900 she is reasonably priced for her actual level, but the client should expect to pair her with oversight rather than hand her an autonomous bookkeeping seat.
Experience
3 years as an accounting assistant at a Cebu BPO, supporting two US clients under a senior accountant. My work was data entry into QuickBooks, categorising transactions, chasing receipts from the client's team, and preparing the bank reconciliation for the senior to review and sign off. I have a BS in Accountancy from the University of San Carlos and I am taking the CPA board exam next year.
Scenario answer
When there is a discrepancy during month-end close, I would first double-check my own work, because most of the time the error is on my side. I would go back through the transactions I entered for the month and compare them line by line against the bank statement to make sure each one matches in both amount and date. If I do not find it that way, I would check whether any transactions are duplicated or missing. Sometimes a transaction is entered twice by mistake, or a deposit was not recorded at all. I would also check whether there are any outstanding cheques or deposits in transit that have not cleared the bank yet, since those would explain a difference without being an error. If I still cannot find it after that, I would raise it with the senior accountant and show her what I have already checked, so she does not repeat my steps. I am careful and I would not want to close the month with a difference I could not explain.
Work sample
CHART OF ACCOUNTS + MONTHLY CLOSE CHECKLIST Tools: QuickBooks Online for the accounting, Dext for receipt capture, Excel for any supporting schedules. CHART OF ACCOUNTS Assets: cash in bank, accounts receivable, inventory, prepaid expenses. Liabilities: accounts payable, credit card payable, sales tax payable, loans. Equity: owner's capital, owner's draws, retained earnings. Income: sales, shipping income, less sales discounts and returns. Cost of goods sold: purchases, freight in, merchant fees. Expenses: advertising, software subscriptions, contractors, salaries and wages, office expenses, bank charges, professional fees. MONTHLY CLOSE CHECKLIST 1. Download the bank and credit card statements for the month. 2. Make sure all transactions for the month have been entered into QuickBooks. 3. Categorise any uncategorised transactions. 4. Match the Shopify and Stripe payouts to the deposits in the bank. 5. Reconcile the bank account and the credit card account. 6. Record depreciation if applicable. 7. Review the profit and loss statement and the balance sheet. 8. Give the reports to the client or the senior accountant for review. The most important check is the bank reconciliation, because if the bank does not reconcile then the financial statements will not be accurate.
This is the artifact every search delivers
Post the role and the same screener runs on your real applicants — the questions arrive already written, and you read every finalist in full before any fee is due.