Resources · stack teardown

What we'd do with a 3-outlet café group, today.

The operation, the inherited mess, and the redesign — the way we approach this kind of business, step by step.

Three café outlets. Singapore. About S$2.4M annual revenue. Inherited a different POS at each location. Books two months behind. Owner does the P&L in Excel.

The operator is technically competent — runs the kitchen, manages staff, deals with suppliers, leases, the lot. What's actually broken is the information layer. They can't answer simple questions in less than an hour: which outlet is profitable this month, what's the real COGS by channel, how much of last week's revenue actually arrived in the bank.

The accounting firm they've been with does monthly bookkeeping and tax. Nothing more. They are not the problem — they're doing what they were hired to do. The gap is everything around them.

Current state · target state

Where the stack is today, and where we'd take it.

Today

Inherited mess

Accounting
QuickBooks Online, manually updated by part-time bookkeeper every fortnight.
Banking
DBS for two outlets, OCBC for the third (legacy).
POS
Different POS at each outlet — inherited from previous tenants.
Receipts & invoices
WhatsApp groups, photos on managers' phones, paper folders.
Payroll
Excel sheet maintained by the operations manager.
Inventory
End-of-month stock-take, written on paper, retyped later.
Dashboards
None. Owner builds a manual P&L in Excel after each month-end.
Month-end close
12 to 16 days. Usually finishes mid-month for the previous month.

After redesign

The Acuitas stack

Accounting
Xero — one ledger across all outlets. Bank feeds wired to DBS and OCBC. Chart of accounts redesigned around outlet × category.
Banking
Aspire layered on top for expense cards, claims, vendor payments, and approval flows — native sync into Xero.
POS
Standardised one POS across outlets. Daily sales export into Xero by outlet and channel. Delivery platforms reconciled separately.
Receipts & invoices
The Receipt Catcher reads from WhatsApp, email, photo. Auto-classified, attached to Xero. Human reviewer confirms low-confidence matches.
Payroll
Connected payroll platform. Manpower cost by outlet, role, and shift visible in dashboard.
Inventory
Inventory management connected to POS and Xero. Stock movements feed COGS daily, not monthly.
Dashboards
Owner dashboard live: cash, revenue by outlet/channel, COGS, labour, margin, exceptions. Weekly summary written by the Compass.
Month-end close
6 days, owned by the Close Captain. Days 1–4 are AI + reviewer; days 5–6 are qualified accountant sign-off.

Phasing

We don't flip everything at once.

For a 3-outlet group, this is roughly 12–14 weeks of design and rollout. The exact pace flexes around your operating calendar — we don't cut over a POS in the week of a delivery platform promotion.

Phase 1 — Foundation (weeks 1–4)

Xero migration from QuickBooks
Chart of accounts redesign by outlet × category
Bank feeds wired
Aspire account opened, cards issued
Receipt Catcher configured, WhatsApp + email channels live

Phase 2 — Operating layer (weeks 4–10)

POS standardisation rolled out across outlets
Delivery platform reconciliation rules
Payroll platform connected
Vendor Chaser, Categoriser, Outlier configured
First full month-end close on the new stack

Phase 3 — Visibility (weeks 10–14)

Owner dashboard live with real data
Weekly summary by the Compass goes to owner
Outlet P&Ls produced monthly
Inventory integration if operation is ready (else deferred)

Trade-offs

The things this redesign costs you.

Most accounting-firm pitches skip this section. We won't — if you're going to commit to a stack redesign, you should know the friction up front.

Standardising POS costs money up front

Running three different POS systems is the silent killer in multi-outlet F&B accounting — every reconciliation breaks differently. We'll usually recommend consolidating to one. That's a hardware and training cost in month one. The payback shows up in month three.

Aspire isn't free, and it adds approval friction

Layering Aspire on top of normal banking adds a card-issuing and management fee. It also means managers can't just buy things on the company DBS card and submit receipts later — they go through Aspire approval flows. Some managers don't love that on day one. After a quarter, they do.

AI agents don't replace your bookkeeper — they change what they do

Our setup doesn't fire the bookkeeper. It changes what they spend time on: less data entry, more review. If your current bookkeeper is happy doing data entry and doesn't want to operate a stack, this isn't a fit for them. Be honest about that early.

Inventory connection only pays off if discipline is there

Connecting inventory to POS to Xero is powerful, but it requires the kitchen to log waste, the manager to do mid-week stock checks, and the team to scan deliveries. If the operation isn't ready for that discipline, we phase it later rather than ship a broken integration.

What the operator gets back

Information layer that catches up with the operation.

Month-end close goes from ~14 days to ~6. The owner sees cash, revenue, costs, and exceptions in their morning dashboard. Managers stop forwarding receipts in WhatsApp groups. The underperforming outlet gets surfaced before three more months pass.

Where the biggest win lands depends on the operation — sometimes COGS visibility, sometimes labour cost, sometimes just stopping a quiet cash leak.

If this sounds like your operation

We'll write a teardown for you — free, before any engagement starts.

Tell us your outlets, banks, POS, and current accountant arrangement. We'll come back with a teardown like the one above for your business — current state, target state, honest trade-offs.

Request a teardown See the live demo