The stack nobody chose on purpose.
It starts innocently. You open the business and buy accounting software, because everyone says to — QuickBooks or Xero, $20 to $30 a month. Sensible. Six months later you need a proper till, so you add a POS — Square, Lightspeed, whatever the hardware vendor suggested — another $49 to $139 per location. Then stock control gets serious and a dedicated inventory app arrives at $39 or more.
Now the three systems disagree with each other daily, so you buy the fourth thing: a connector to sync sales into the books, $20 a month and worth every rupee of the pain it half-solves. At no point did anyone approve a $130-to-$300-a-month software budget. It assembled itself, subscription by reasonable subscription — and it renews forever.
The receipt, itemised at list prices.
Here is what the standard stack costs a single-location product business, using each vendor’s published list price for the lowest comparable plan, checked August 2026. Your mix will differ; the shape of the bill will not.
And the fifth line is the one most owners eventually add: once the systems disagree often enough, a bookkeeping service takes over reconciling them — Bench from roughly $190 a month, Zeni from $549. That is how a $138 stack quietly becomes a $327+ stack: the service exists mostly to referee the other four subscriptions.
- Accounting (QuickBooks Online): $30/month
- Point of sale (Square for Retail): $49/month per location
- Inventory control (Zoho Inventory): $39/month
- POS-to-books connector: $20/month
- Subtotal, software only: $138/month — four logins, four invoices
- Add a bookkeeping service to referee it all: from $189/month → $327+/month total
The costs that never appear on the receipt.
The subscriptions are the visible half. The invisible half is time: every month, someone reconciles the POS report against the accounting software against the inventory app, and they never quite agree. A refund processed on the till but mapped wrong by the connector. A stock adjustment that exists in one system only. Card fees that land as a lump the books cannot explain. Five hours a month of a capable person’s time is a conservative estimate — price that at any wage and it dwarfs the software bill.
Then there is the price creep. Stitched stacks compound it four ways at once: QuickBooks Plus went from $90 to $115 a month inside two years, POS vendors charge per location as you grow, connectors meter by order volume. Four vendors, four annual increases, and cancelling any one of them breaks the chain the other three depend on. That dependency is precisely why the increases stick.
“Integration” is the most expensive word in software.
The connector’s job sounds simple: copy sales from the till into the books. But a real trading day is not simple. Split payments, partial refunds, exchanges, loyalty discounts, voided lines, cash drawer corrections — each one has to survive a mapping between two systems that model the world differently. The connector translates; translations drift; and the person who notices the drift is you, at month end.
This is the structural truth the stack cannot escape: when the till, the stock, and the ledger are three databases, agreement between them is a maintenance job. Not because any vendor is bad — because reconciliation is what separate systems require. You are not paying four subscriptions for software. You are paying them for the privilege of being the integration layer yourself.
What changes when it is one ledger.
Full disclosure: Nonari is the product we build, and this section is the reason we built it. When the point of sale, the inventory, and the accounting are one double-entry ledger, a sale at the till IS the accounting entry. Stock decreases, cash increases, revenue and cost of goods post — one event, one record, the same second. There is no sync to run, no mapping to drift, no month-end argument between systems, because there are no systems to disagree.
The arithmetic follows from the architecture: one ledger means one subscription — $29 a month for a single location, $70 for up to three branches, every feature on both plans, unlimited users. Against the $138 software stack it is roughly a fifth of the cost; against the $327 stack-plus-service it is under a tenth. The AI bookkeeper that drafts entries from supplier invoices and bank lines is included, because on one ledger it has the whole picture to draft from.
- One event, one record: a sale posts to stock, cash, revenue, and COGS instantly
- Zero connectors, zero mappings, zero reconciliation between systems
- One renewal to watch instead of four compounding ones
- $29/month (one location) or $70/month (three branches) — 15-day trial, no card
When the stack is actually the right answer.
Honesty clause: sometimes the stitched stack earns its cost. If your accountant runs a QuickBooks-only practice and does your tax filing inside it, the ecosystem is worth real money. If your restaurant depends on a specialist POS’s kitchen features, or your US payroll and benefits run through a dedicated provider, those tools justify themselves.
The stack is also right when any single system is genuinely load-bearing at enterprise depth — a warehouse running wave picking, a chain on NetSuite-grade consolidation. The one-ledger argument is strongest exactly where most product SMBs actually live: one to a handful of locations, standard retail or restaurant operations, and a monthly software bill that grew past what the business ever decided to spend.
Audit your own stack in ten minutes.
Pull up your card statement and answer four questions. One: how many back-office subscriptions do you pay — books, till, stock, connectors, reporting add-ons? Two: what is the true monthly total, including per-location and per-order fees? Three: how many hours a month does someone spend making the systems agree, and what is that person’s time worth? Four: when did each vendor last raise its price?
Most owners who run this exercise find a number between $130 and $500 a month plus five to fifteen hours of reconciliation. Whatever you decide to do about it — consolidate onto one ledger, renegotiate, or keep the stack deliberately — decide it on the real number. The stack assembled itself without a decision; the least it owes you is one honest look at the receipt.