A close held together
by manual handoffs.
Example Co. is a 35-person, owner-operated industrial-supplies distributor in the Carolinas, with a finance team of four and a nine-day close. It’s constructed, so the numbers teach the method; they don’t prove results.
- ILLUSTRATIVE TEAM
- 35 people
- MONTH-END CLOSE
- 9 business days
- MANUAL WORK IDENTIFIED
- 61 hours/month
Some workflows need a design.
Others need a reason to wait.
Three candidates to validate, in order: the reporting pack, then reconciliations, then budget-vs-actual drafts. Build only after validation and an agreed valuable use for the released capacity. The board & lender packet waits for Phase 2. Contract renewals don’t need software; they need an owner.
These are three roadmap recommendations, not three included designs. The standard audit includes two workflow designs, selected from the ranked recommendations. This excerpt does not assign the two design deliverables; implementation is scoped separately.
The reason beside the recommendation.
These are illustrative sequencing recommendations, not approval to spend. Proceed only after checking source data, review effort, implementation and ongoing costs, and an owner’s specific use for the released capacity. If the decision depends on cash savings alone, this sample does not support the proposed build. Use existing controls where they solve the problem; defer work whose dependencies are not ready.
Monthly reporting pack
Extracts from the ERP and the BI tool are copied into a spreadsheet chain and formatted by hand every month. The design validates each extraction at source and automates the compilation; the controller reviews the pack before it leaves the building. Nothing writes to the ledger.
22 baseline hours/month × 70% (report-compilation class) = 15.4 modeled recovered hours/month.
15.4 × $75 × 12 = $13,860 modeled annual capacity value.
Account reconciliations & tie-outs
Inventory, intercompany and accrual tie-outs are worked line by line. The design matches exceptions-first and attaches the source rows to every exception. A parallel run and agreed tolerances come before the manual checklist changes.
15 baseline hours/month × 60% (reconciliation class) = 9.0 modeled recovered hours/month.
9.0 × $75 × 12 = $8,100 modeled annual capacity value.
Budget-vs-actual department packs
Variance commentary is drafted from the model; department owners approve their own lines. The draft saves the writing, not the judgment.
10 baseline hours/month × 50% (drafting class) = 5.0 modeled recovered hours/month.
5.0 × $75 × 12 = $4,500 modeled annual capacity value.
Board & lender packet
No benefit is counted in the 90-day model. The packet depends on the reporting pack’s definitions. Building it first would automate a disagreement.
8 baseline hours/month · no benefit modeled.
Contract & subscription renewals
Assign an owner and turn on the existing system’s renewal alerts. Six hours a month doesn’t justify a build; it justifies a name on a calendar. No benefit is modeled.
6 baseline hours/month · no benefit modeled.
Capacity has value.
It is not cash.
29.4 hours × $75 × 12
$26,460
29.4 = 15.4 + 9.0 + 5.0 modeled recovered hours per month from the three implementation candidates. The $75 loaded hourly rate is the composite’s assumption: one blended team rate, loaded. This is time available for other work, not a payroll saving or an achieved result.
This model counts three candidates at workflow-specific capture assumptions. The homepage screen applies one blended assumption to all five workflows.
The hypothetical cash case, on its own lines.
- Cancel one overlapping reporting subscription (hypothetical; verify the contract and date)
- +$2,400/yr
- New recurring tooling: business AI seats (≈$900) and an automation platform (≈$960)
- −$1,860/yr
- Modeled net recurring cash change
- +$540/yr
- Implementation, if Unwrapped builds
- $30,000
- Simple cash payback on the $30,000 build
- 55.6 years
So the readout asks one question.
At an assumed $540 a year in net recurring cash savings, recovering the $30,000 implementation takes about 55.6 years. Including the $7,500 audit makes that about 69.4 years before any credit. If the full audit fee has been paid and a follow-on SOW is signed within 30 days of the readout, the one-time $3,750 credit reduces total upfront cost to $33,750 and simple cash payback to 62.5 years. No cash saving in this sample has been realized. A cancellation contract supports an estimate; payment records and actual reduced outflows, compared with the documented baseline including incremental costs, would establish a realized result. This simple calculation assumes unchanged savings and costs; it excludes ramp time, financing, tax and the time value of money. The case is capacity, which is why the readout asks what the 29.4 hours are for.
The modeled monthly capacity value is $2,205: 29.4 hours × $75.
In this sample, an intermediate two-day reduction would move the close from nine to about seven days. The Phase 1 target is six days, a three-day reduction from the baseline, and the planned staff-accountant backfill can be re-evaluated in Q1. That is a decision for the owner, not a promised saving.
The reporting pack, with a review gate.
A sketch of the sample design, not a working automation or a client result. These are checks to implement; the audit reviews source assumptions and design logic, not production performance.
- Collect the agreed ERP and BI extracts for the reporting period.
- Validate source totals and route mismatches to the controller.
- Compile the pack only after the source checks pass.
- The controller reviews and approves the pack before distribution. If checks fail, retain the manual process.
Recovery rates of 70%, 60% and 50% above are illustrative assumptions, not measured outcomes. Confirm them during separately scoped implementation through parallel runs that include exceptions and review time before production acceptance.
Every tool gets a verdict.
Five of the twelve tools the five workflows touch. Tools outside the audit’s scope are listed in the full package as not assessed.
- ERPSystem of record. Extraction is validated at source.
- Keep
- BI tool, licensed but lightly usedBecomes the reporting pack’s single source.
- Augment
- Reporting add-in overlapping the BI tool$2,400/yr. Verify the contract and cancellation date first.
- Retire candidate
- Legacy report-writer$0. Unused; remove it from the stack.
- Retire candidate
- Free personal AI chat accountsMove to business AI seats with an approved-tool list and a review rule.
- Replace
The roadmap starts with a light governance step: an approved-tool list and a review rule.
Your evidence. Your decisions.
The engagement applies the same method to your five workflows with your own numbers: the tool stack audit, the leverage map, the 90-day roadmap, two workflow designs and the Working ROI Model. This excerpt shows the structure; the scoring weights stay in the engagement.
Unwrapped inventories, maps, validates, reconciles and orchestrates. It does not move money, post to the general ledger, contact your customers, build chatbots or act as an ERP implementer. Each design names its review points and a checker who did not write the prompt.
THE LETTER · BY PAUL MATTSON
Dated. Signed.
Every figure sourced.
One finding about AI in owner-led companies, weekly, with the number that held and the number that didn’t.