How Watchbaze sizes every restock before the money leaves the account

Watchbaze sells watches online through Shopify. Like most shops at that size, the hard part was never the selling — it was the ordering. Every restock ties up a five-figure amount weeks before any of it comes back as revenue, and the only number available at the moment of deciding was the bank balance.

A bank balance is a bad number to order against. It shows what is in the account today. It does not show the supplier invoice due in eleven days, the VAT payment at the end of the month, or the payout from the shop that has not landed yet. Order against it and you either commit too much and get tight, or — far more often — you order too small and run out of the products that were selling.

What they connected

Setup was accounting first, shop second. That order matters: accounting bookings arrive already assigned to an account, so nothing has to be categorized by hand, and the plan ends up using the same numbers the tax advisor does.

• Accounting — actuals, accruals and open items, including the supplier invoices that were not due yet.
• Shopify — orders, average order value and refunds, which became the drivers under the revenue line.
• Payments and ads — payouts, fees and ad spend, so the gap between spending on an ad and collecting the order it produced is visible instead of assumed.

The first usable forecast was on screen in about twenty minutes. Not a model anyone built — the accounting structure was already the structure of the plan.

How the restock is planned now

The restock is not typed in as a number any more. It is a scheduled line with a frequency, a day of the month and an average purchase value driver behind it, so the spend scales with the revenue forecast instead of being re-estimated each cycle. Because Foreqast works day by day, the order lands in the cash curve on the date the invoice will actually hit, not smeared across a month.

What that changes at the moment of deciding:

We can see what a restock does to our account before we order it — not once the invoice arrives.

The second effect: ad budget

The restock was the reason they started. The ad budget is where the model earned its keep a second time. Revenue follows ad spend, and COGS, shipping and payment fees follow revenue — so raising the budget moves the whole chain at once, including the cash gap in between, where the spend has already gone out and the orders it bought have not been paid out yet.

Instead of capping the budget at a number that felt safe, Watchbaze can raise it in the simulator, look at the dip, and check that the trough still clears the restock due a few weeks later.

Where they are now

Restock cycles are planned ahead rather than reacted to, the ad budget is set against a visible floor, and the monthly spreadsheet rebuild is gone. The forecast updates itself on every sync.