Acme Corporation
Finance Function Assessment

02How to read this assessment

The lens

Till evaluates a finance function across eleven topics. This document groups those eleven topics into three lanes, because the lanes are where the fixes live.

People asks who holds the knowledge, who makes the decisions, and whether the function survives a resignation. Process asks whether the work that happens every month is defined, repeatable, reviewed, and forward-looking. Technology asks whether the systems remove work or create it.

The order matters. Technology amplifies whatever process it is pointed at. Process only holds if the right people own it. So we diagnose in the order people, process, technology, and we prescribe fixes in the same order.

The maturity scale

Each topic is scored today and given a twelve-month target.

A target of 3 is not a low bar. It means the function no longer depends on any one person. For Acme, that is the single most valuable move available.

03Company snapshot

04Scorecard

Three topics are Critical and they share one root. The Controller is the close, the close is manual, and revenue enters the close by re-keying. Fix the root and the Critical column clears together.

05People

5.1 Org structure and rolesSTATED

Current state. Four people run finance for a $29M business, and the shape of the team is reasonable for the size. The problem is not headcount. It is that the Controller performs every task that requires judgment: bank reconciliations, accruals, the export and reclassification of revenue from the field-service platform, the covenant calculation for the bank, and every one of roughly 1,300 manual journal entries a year. The AR specialist, the AP clerk, and the office manager each run a transactional lane. None of them can close the books. The outside CPA firm prepares the tax return and performs an annual review; it plays no role in monthly reporting.

Gap. There is no documentation of the close, no cross-training, and no second person who has ever run it. The Controller's twelve years of tenure are the reason the books are clean and the reason the risk is invisible. A two-week notice would leave Acme unable to produce a month-end close, a covenant certificate, or a bank reconciliation until a replacement reverse-engineered the work.

Impact. This is the finding that turns every other finding into a deadline. In our experience across this revenue band, the most common trigger for a finance crisis is not a bad quarter. It is the sole accountant resigning. Acme's growth rate makes it more likely, not less, because the Controller's workload grows with every invoice and every entity.

What good looks like. The close is documented step by step, with a checklist that a second person runs every quarter to prove the documentation works. The AR specialist owns revenue reconciliation with the Controller reviewing, not performing. The Controller's time shifts from producing the close to reviewing it and toward the forward-looking work in Section 6. Score today: 2. Target: 3 within six months.

5.2 The CEO's altitudeSTATED

Current state. The CEO reviews transaction-level detail at month end and approves every payment over $10K by email. On the call, this was described as "staying close to the numbers." It is more accurately described as compensating for the absence of a report. The CEO does not have a management pack, a margin-by-contract view, or a cash forecast, so the only way to feel confident is to read the ledger.

Gap. The CEO is operating one level below the seat. Time spent reviewing transactions is time not spent on the $50M plan, the second acquisition, or the pricing escalators that are being left on the table (Section 6.2). The email approval of payments is a control in name only; it happens after the invoice is entered and before anyone has checked it against a purchase commitment.

What good looks like. The CEO consumes a monthly pack that answers the questions the ledger review is trying to answer, reviews a 13-week cash view every other week, and approves spend through a documented matrix rather than an inbox. This is a process fix that changes a people behavior, which is why it appears in both sections.

5.3 Strategic financeNOT EXPLORED

Current state. The CEO wants to reach $50M in four years and is evaluating a second acquisition. The first acquisition closed in March. There is no acquisition model for either transaction, no integration plan for the first, and no one on the team whose role includes forward-looking finance. The outside CPA was not involved in the first deal beyond tax structuring.

Gap. Acme has a growth strategy and no finance function to test it. The $50M plan has not been modeled. The first acquisition has not been measured against its own thesis because its results sit on a separate ledger. The second acquisition will be evaluated the same way the first was: on the owner's judgment, without a model.

Impact. At $29M with a $2.0M line of credit and cash that swings from $600K to $2.8M in-year, a second acquisition without a cash model is a bet on timing. The judgment may be right. The business cannot currently show that it is.

What good looks like. A three-year operating model tied to the budget, an acquisition template that can be run on any target in a week, and a post-close integration checklist that starts with the general ledger. Score today: 1. Target: 3.

5.4 Risk managementINFERRED

Current state. Insurance is reviewed at renewal only. There is no cyber coverage. Sales tax on services in two states is unresolved. Customer concentration is moderate (top 10 at 44%, largest at 11%) but not monitored. The bank covenant is computed quarterly by the Controller with no second check.

Gap. Each item is small. Together they describe a company that manages risk when a renewal notice or a bank letter forces it to. The sales tax question is the one with a number attached; two states of unpaid service tax on a $29M base is a real liability that a buyer or lender would price.

What good looks like. A one-page risk register reviewed quarterly, a nexus study on the two open states, and the covenant calculation performed and reviewed by two people. Score today: 2. Target: 3.

06Process

6.1 Accounting operations and the closeCONTRADICTED

Current state. On the call, the close was described as taking "about a week." The general ledger shows the books closed on business day 18, 19, 22, 20, 21, and 19 across the last six months. The gap between the answer and the data is not a matter of honesty. The CEO does not see the close; the CEO sees the moment the P&L arrives and, from that vantage, a week feels right.

The close itself is complete and accurate when it lands. The Controller reconciles every bank account, books accruals, exports revenue from the field-service platform, reclassifies it into the right accounts, and types the Gusto payroll summary into QuickBooks by hand. Every one of those steps is manual, and each waits on the one before it.

Gap. Three weeks is too long for a business growing 18% a year, and the length is structural, not personal. The close waits on re-keyed revenue, re-keyed payroll, and one person's calendar. There is no close checklist, no cut-off procedure for the field-service platform, and no reviewer.

Impact. Every decision the CEO makes in the first three weeks of a month is made on the prior month's numbers. Margin drift is discovered a quarter late. The covenant is tested on a close that finishes after the certificate is due, which means the certificate is being prepared from preliminary numbers.

What good looks like. Books closed by business day 7 with a documented checklist, revenue and payroll entering the ledger without re-keying, and a reviewer who signs the close. Score today: 2. Target: 4.

6.2 Revenue recognition and billingINFERRED

Current state. Revenue is recorded when an invoice is raised in QuickBooks. Managed-service contracts are billed monthly, quarterly, or annually in advance; the annual and quarterly prepayments, roughly $1.4M a year, are recognized as revenue in the month invoiced. There is no deferred revenue account. Project revenue is billed on milestones; work performed but not yet billed lives in the field-service platform and never reaches the balance sheet. At FY2025 year-end that unbilled work was about $680K.

Billing runs through the AR specialist, who re-keys work orders from the field-service platform into QuickBooks invoices. About 1,850 invoices a year pass through this step. Roughly 3% are corrected after a customer dispute, which is the only error detection in the process. Of the 62 managed-service contracts, 35% carry annual price escalators, and those escalators have been applied inconsistently.

Gap. Three distinct problems share this section. First, the income statement is wrong in both directions: prepayments pull revenue forward, unbilled work pushes it back, and the two do not net to zero in any given month. Second, revenue leaks at the re-keying step, and the leak is detected by customers rather than by Acme. Third, contract escalators that Acme has already negotiated are not being billed.

Impact. The monthly gross margin the CEO is worried about is partly noise from timing, which means the real margin trend is unknown. The escalator gap is unbilled revenue at 100% margin; on a 62-contract base with 35% carrying escalators, it is worth quantifying in the first two weeks of any engagement. A lender or buyer performing diligence would restate revenue before anything else.

What good looks like. A revenue recognition policy for both contract types, a deferred revenue schedule that rolls forward monthly, unbilled work-in-progress on the balance sheet, invoices generated from the field-service platform without re-keying, and an escalator calendar that bills every contract on its anniversary. Score today: 1. Target: 4. This is the section that most changes what Acme's financials say.

6.3 Financial reportingSTATED

Current state. Each month the P&L is exported from QuickBooks to Excel. There is no departmental P&L, no contract-level margin, no balance sheet commentary, and no management pack. The bank receives a quarterly covenant package. The CEO is the only reader of the monthly output and, as noted, reads the ledger instead.

Gap. Acme produces financial statements but not management reporting. The difference is that statements say what happened and reporting says why. The question the CEO asked on the call, "why is margin down four points and which contracts did it," cannot be answered from anything currently produced.

What good looks like. A monthly pack, delivered by business day 10, with a one-page summary, P&L against budget with variance commentary, margin by service line and by top-20 contract, balance sheet and cash summary, and a short list of decisions needed. Score today: 2. Target: 4.

6.4 Forecasting and budgetingSTATED

Current state. The CEO and Controller build an annual budget in December. It is not re-forecast during the year and variance is not reviewed monthly. There is no 13-week cash forecast. Cash decisions are made from the bank balance.

Gap. The budget is a document, not a process. Six months in, it describes a company that no longer exists. The absence of a cash forecast is the more urgent gap; the line of credit was drawn to $1.2M at the Q2 peak to cover payroll timing, which is what a business does when it discovers a cash need instead of anticipating it.

Impact. Acme is running an 18% growth year, an acquisition integration, and a $50M plan without a forward view of any of them.

What good looks like. A 13-week cash forecast updated weekly, a rolling twelve-month forecast updated quarterly, and a monthly variance review that takes thirty minutes because the pack in 6.3 already explains the numbers. Score today: 1. Target: 3.

6.5 Cash management and treasurySTATED

Current state. Cash swings between $600K and $2.8M inside a year. DSO is 61 days against net-30 terms. DPO is 24 days; the AP clerk pays bills on receipt through bill.com once the Controller approves. The line of credit is used reactively.

Gap. Acme pays its vendors in 24 days and collects from customers in 61. On $29M of revenue, every day of DSO is roughly $80K of cash. Bringing DSO to 45 days would release on the order of $1.3M, more than the peak line draw. The working capital problem is a collections cadence problem, and nobody owns collections.

What good looks like. A weekly AR review with named owners for every balance over 45 days, payment terms managed to a DPO target rather than paid on receipt, and the line of credit drawn against a forecast rather than a surprise. Score today: 2. Target: 4.

6.6 Compliance and controlsINFERRED

Current state. The Controller creates and posts every journal entry, about 1,300 a year, with no second reviewer. Bank reconciliations are performed but not reviewed. Payroll is run by the office manager with no second check. All four QuickBooks users hold admin rights. The vendor master has duplicates. There is no documented approval matrix; the CEO's $10K email approval is the only spend control.

Gap. Acme has no segregation of duties anywhere in finance. This is common at this size and it is not an accusation. It is a description of a control environment that relies entirely on the integrity and accuracy of individuals, which has worked because the individuals are good. A lender's field exam, an acquirer's diligence, or a single error in a payroll run would each expose it.

What good looks like. An approval matrix by spend type and amount, journal entries reviewed by someone other than the preparer, payroll reviewed before submission, QuickBooks roles restricted to what each person does, and a vendor master cleanup with a new-vendor approval step. Score today: 1. Target: 3.

6.7 KPI and performance trackingNOT EXPLORED

Current state. Technician utilization, about 71%, is computed quarterly in Excel by operations. Backlog is visible in the field-service platform. Contract margin is not tracked. There is no KPI cadence and no shared definition of the metrics that would run this business.

Gap. A technical services company runs on four numbers: utilization, gross margin by contract, backlog coverage, and DSO. Acme has one of them, quarterly, in a spreadsheet owned by another department.

What good looks like. A weekly scorecard with those four numbers and named owners, sourced from the field-service platform and the ledger without manual assembly. Score today: 1. Target: 3.

07Technology

7.1 Systems mapSTATED

The picture is typical for the size. Each system is adequate on its own. The only working integrations are the ones the vendors built for free (bill.com and the bank feed). Everything that matters to Acme's economics, which is the field-service platform, enters the ledger through a person.

7.2 Where the work is created

Three re-keying points generate most of the manual effort and most of the error in the finance function.

Field-service platform to QuickBooks, revenue. About 1,850 invoices a year are typed twice. This is where the 3% dispute rate originates and where unbilled work-in-progress falls out of the financials.

Gusto to QuickBooks, payroll. A summary journal entry typed monthly. Low error risk, but it sits on the close's critical path and is invisible to anyone but the Controller.

QuickBooks to Excel, reporting. Every report, the budget, the covenant calculation, and the KPI sheet start with an export. The exports are the reason the reporting in Section 6.3 is a P&L and not a pack: assembling anything richer by hand takes longer than the month.

7.3 What stays, what changes, what gets added

Stays. QuickBooks Online can carry a $29M single-entity business if the chart of accounts is rebuilt around service lines and a deferred revenue structure is added. Acme does not need a new general ledger this year. It needs a better one inside the ledger it has. bill.com stays. Gusto stays.

Changes. The field-service platform becomes the source of invoices and of work-in-progress, flowing into QuickBooks rather than being retyped. The outcome is invoices that match work orders, unbilled work visible on the balance sheet, and a close that no longer waits on the AR specialist's keyboard. The specific integration path is validated in the first weeks of an engagement, once we have hands on both systems; we prescribe the outcome here, not the mechanism.

Added. Two layers. First, a spend platform with corporate cards, approval workflows, and receipt capture, which gives Acme the approval matrix from Section 6.6 as a system rather than a memo. Second, a planning and reporting layer that reads the ledger directly, so the pack, the forecast, and the scorecard are refreshed rather than rebuilt.

Deferred. Consolidation of the second entity is handled first in process (a common chart of accounts and a monthly consolidating workbook), then in technology once both entities close cleanly. Moving to a multi-entity general ledger is a decision for month nine or later, after the second acquisition question is settled. Score today: 2. Target: 4.

08Cross-cutting findings

Four findings explain most of the symptoms above.

The close is single-threaded. One person, no documentation, no reviewer, three weeks. Every other timeline in this document is gated by it.

Revenue leaks between the field-service platform and the ledger. Prepayments pulled forward, unbilled work left off, escalators unbilled, invoices retyped. The income statement is a good approximation with a wide error band, and the margin drift the CEO is worried about cannot be diagnosed until the band narrows.

Nothing looks forward. No cash forecast, no re-forecast, no model for the $50M plan or the acquisitions. The line of credit is the forecast.

The second entity was bought, not integrated. Separate ledger, separate chart, separate bank accounts, six months in. Its results are invisible inside Acme's, and the second acquisition would repeat the pattern.

The first finding is a people problem. The second and third are process problems. The fourth is both. Technology appears in none of them as a cause, and in all of them as part of the fix, which is the point of doing this assessment in the order we did.

09Recommended path

The sequence is stabilize, build, amplify, measure. Each phase names what changes in people, process, and technology, and what Acme's team has to enable. No phase depends on new software except where stated.

Days 1 to 30: stabilize

People. Document the close as the Controller performs it. Cross-train the AR specialist on bank reconciliations and revenue reconciliation. Shift the CEO's payment approval from email to a written matrix.

Process. Establish a close checklist with owners and a target of business day 12 for this phase. Quantify the escalator gap and the unbilled work-in-progress. Draft the revenue recognition policy. Start a 13-week cash forecast in a spreadsheet; the forecast matters more than the tool.

Technology. Restrict QuickBooks roles. Clean the vendor master. No new systems.

Acme enables. Access to both QuickBooks files, the field-service platform, bill.com, Gusto, and the bank portal in week one. Two hours a week of the Controller's time for documentation.

Days 31 to 90: build

People. The Controller moves from preparing the close to reviewing it. Collections gets a named owner and a weekly cadence. The CEO stops reviewing transactions because there is now something better to read.

Process. Rebuild the chart of accounts around service lines and add deferred revenue and unbilled WIP. Book the opening balances for both. Apply the revenue policy. Deliver the first monthly pack by business day 10. Set the approval matrix live. Bill every missed escalator.

Technology. Connect the field-service platform to QuickBooks for invoicing and work-in-progress. Stand up the spend platform with the approval matrix built in. Payroll enters the ledger by integration.

Acme enables. A decision on the field-service platform integration path by day 45. Operations time to define the four KPIs and where each is sourced.

Months 4 to 6: amplify

People. Quarterly cross-training test: someone other than the Controller runs the close from the documentation. The CEO reviews the 13-week cash view every other week and the pack monthly.

Process. Close by business day 7. Rolling twelve-month forecast replaces the static budget. Weekly scorecard with utilization, contract margin, backlog coverage, and DSO. Consolidating workbook for the second entity on the common chart of accounts. Nexus study on the two sales tax states.

Technology. Planning and reporting layer reads the ledger directly; the pack and scorecard refresh without exports.

Acme enables. A DSO target and the authority for the collections owner to enforce terms.

Months 7 to 12: measure and decide

People. Risk register reviewed quarterly. Covenant calculation prepared and reviewed by two people.

Process. Three-year operating model for the $50M plan. Acquisition model template, run first on the March acquisition to measure it against its thesis, then on any second target.

Technology. Decide on multi-entity consolidation in the general ledger based on whether the second acquisition proceeds.

Acme enables. The strategic questions. This is where the CEO's time goes once it is no longer spent in the ledger.

By month twelve, every topic on the scorecard reaches its target. The finance function documents itself, closes in a week, reports margin by contract, forecasts cash thirteen weeks out, and survives a resignation. That is the infrastructure a future finance leader inherits and directs, rather than rebuilds.

10What we could not verify

This section exists in every Till assessment. It lists what the analysis rests on and what would change it. "No contradiction between our findings" is not the same as "no risk."

11Appendix

A. Eleven topics mapped to the three lanes

B. Evidence tags

C. How a live assessment is built

A Till assessment draws on three sources. A recorded assessment call of about ninety minutes with the decision maker and the finance lead. General ledger exports for the trailing twenty-four months: profit and loss, balance sheet, trial balance, and the full general ledger detail, analyzed line by line rather than summarized. A scouting report on the business, its market, and its buying context. The financial analysis is performed independently of the call so that stated claims can be tested against the data; that is where CONTRADICTED findings come from, and they are usually the most useful thing in the document.

The output is this document, delivered within ten business days of the call, followed by a working session to walk through the findings and the recommended path.