The questions that live between two systems.
Every business already knows what each system says on its own. Nobody is short of a sales report. What is missing is the answer that needs two systems at once, and that is exactly the answer no single tool will ever give you, because it only holds one half.
A blend is not an integration. It is a join.
Two systems can both be connected and still be useless together. Shopify calls it order_id, Xero calls it InvoiceNumber, and the same shoe is BOS-114 in one and 114-BLK-9 in the other. Until something reconciles that, you do not have data. You have two spreadsheets in a trench coat.
Puts them side by side
Two tiles on one screen. Revenue from Shopify, costs from Xero. You do the join in your head, every time, and slightly differently every time. Nobody can audit a join that lives in your head.
Reconciles them once
The SKU map is written down. The fee treatment is written down. The join runs every night whether you are looking or not, and change the rule once and eleven reports move together.
Nine questions worth two systems.
EVA suggests these once she can see what you have connected. Add Deputy on top of Xero and the labour blend appears on its own - you do not go looking for it.
+
What is labour actually costing us, by site, by week?
Deputy holds the hours and the roster. Xero holds the revenue and the on-costs. The percentage only exists when they meet.
Neither system can tell you this alone.
+Which channel actually makes money after fees and returns?
Shopify reports gross sales. It does not know your COGS, your merchant fees as booked, or the credit notes you raised three weeks later.
Shopify will happily show you a number that is wrong by twelve points.
+How much cash is sitting in stock that has not moved in 90 days?
Cin7 knows what is on the shelf and when it last moved. Xero knows what you paid for it and what you still owe on it.
Usually the largest number nobody is looking at.
+What is our WIP and lockup, by partner, this month?
Practice Manager holds time and jobs. Xero holds what was actually billed and actually collected. Lockup is the gap.
The classic firm question. Two systems, always.
+
Which completed jobs are missing a signed document?
The job system says the work is done. FYI says whether the paperwork exists. Billing a job whose documents are not filed is a compliance problem, not an admin one.
Nobody checks this until an audit does.
+Which clients are unprofitable once you count the chasing?
Karbon holds the effort. Xero holds the fee. Some of your favourite clients are your worst clients and the data has always known.
Uncomfortable, and the highest-value question on this page.
+What is our true cost per head, including on-costs?
Salary is the number people quote. Super, leave movement, workcover and payroll tax are the number that hits the P and L.
Usually a fifth higher than the number in the budget.
+Why does the bank never match the sales report?
Stripe settles net of fees on a lag, in batches that do not line up with invoices. Reconciling that by hand is somebody's Tuesday, every Tuesday.
A blend does it nightly and flags only the breaks.
++True margin by product, across every channel.
Three systems. Landed cost from Unleashed, channel price and fees from Shopify, the actual posted result from Xero.
Three-way blends are where this stops being a reporting tool.
One question, four systems, one answer.
A real demo tenant inside VibeCFO - you can open it on a call. The question a real owner asked: “Which of my channels is actually worth expanding?”
+++→ one table nobody could build by hand| Channel | Gross sales | Landed COGS | Channel fees | Labour to serve | True margin |
|---|---|---|---|---|---|
| Wholesale | $1,284,400 | $742,900 | $0 | $61,200 | 37.4% |
| Retail stores | $896,150 | $451,300 | $18,400 | $212,700 | 23.8% |
| Online (Shopify) | $412,880 | $246,100 | $29,930 | $47,400 | 21.2% |
| Marketplace | $203,400 | $128,700 | $38,600 | $21,900 | 6.9% |
Landed COGS unleashed.stock_movements
Channel fees xero.bank_transactions
Labour to serve deputy.timesheets
He had been planning to push harder on marketplace, because it was the fastest-growing line on the sales report. At 6.9 per cent it was growing him toward nothing. He moved the spend to wholesale and left marketplace where it was.
You bring the question. EVA brings the plumbing.
In your own words
"Show me labour as a percentage of revenue by site since July." No table names. No SQL.
She shows her working
Which connections she will use, which join, which definition of revenue. You correct her before anything runs.
It becomes a report
Once it is right, pin it. It runs on the same definition every night, for everyone allowed to see it.
Set a Pulse on it
Labour over 32 per cent at any site and your phone buzzes before Friday, not after month end.
Blending, answered.
What does it mean to blend two connections?
Blending means joining records from two different systems so a single figure can be calculated across both - labour hours from a rostering system against revenue from an accounting ledger, for example. It requires the two systems to be reconciled to each other, not just displayed side by side.
Why can a dashboard tool not do this?
A dashboard puts two tiles on one screen and leaves you to do the join in your head, slightly differently every time. A warehouse reconciles the two once - the SKU map and the fee treatment are written down - and every report that uses them gets the same answer.
How long does it take to set a blend up?
Mapping your SKUs, your client names and your fee treatment is real work and it is most of what the first two weeks are for. It is also why the answer is still right in month nine, when whoever set it up has moved on.
Do I need to know which systems answer my question?
No. You describe the question in your own words and EVA works out which connections it needs, shows you the join she intends to use, and lets you correct her before anything runs.
Bring two systems.
Your ledger and the operational system you actually run on. That pair alone answers questions neither could answer before.