You'll know when it happens
Every brand runs on spreadsheets at the start, and that's fine. A Google Sheet with your SKUs, a tab for cash, a tab for the next PO, a bookkeeper who tidies up Xero once a quarter. It works. It works right up until the week it very suddenly doesn't.
The symptoms are weirdly consistent. You're out of stock on your best seller and you've got a pallet of the thing that stopped selling in March. Someone asks what your gross margin was last month and three people give three different answers. The cash forecast says you're fine and the bank balance says otherwise, because the forecast didn't know about the Meta bill, the 3PL invoice or the fact that half of Q3's revenue is sitting in Shopify Payments and Amazon's 14-day hold. You find a formula in the inventory sheet that's been pointing at the wrong column since someone added a row in February. And the founder spends Sunday night 'fixing the model' instead of, say, sleeping.
It tends to hit somewhere between £1M and £5M in revenue, which isn't a coincidence. That's the point where you've got enough SKUs that nobody can hold them all in their head, enough channels (Shopify, Amazon, wholesale, maybe TikTok Shop) that the numbers no longer live in one place, enough stock on the water that a bad forecast costs real money, and enough spend that a month of wrong numbers is a month of wrong decisions. The spreadsheets didn't get worse. The business got bigger than one person's ability to keep them honest.
The tempting response is a bigger spreadsheet. A better model. A consultant who builds you a 40-tab monster that only they understand. Don't. The fix isn't a smarter sheet; it's moving each job the sheets were doing into software that does that one job properly and talks to the others.
This piece is about the four jobs, the order to move them in, and four tools that are built for brands at exactly this size: Finaloop for the books, Flagship for forecasting demand, GoodDay for running inventory, and Iris for planning the cash.
The four jobs, and the order to fix them in
Strip it back and your spreadsheets were doing four things. They were keeping the books (what happened, in money). They were forecasting demand (what's going to sell). They were managing stock (what's where, what's on order, what's about to run out). And they were planning cash (can we afford the next PO, the next hire, the next Q4). Four different jobs, four different people who care, four different ways to get it wrong.
The order matters because each one feeds the next:
| Job | What the spreadsheet was doing | How it breaks | Fix it with | Why this order |
|---|---|---|---|---|
| 1. Books | Recording revenue, COGS, fees, returns, by channel | Shopify payouts don't match orders, Amazon fees are a mystery, COGS is a guess, month-end takes three weeks | Finaloop | Everything else is built on these numbers. A forecast on bad books is a confident forecast of the wrong thing |
| 2. Demand | Guessing how much of each SKU will sell over the lead time | Based on last month, ignores seasonality, promos and stockouts; one person's gut feel in a cell | Flagship | Once you know what actually sold (books), you can model what's going to sell. This drives the PO |
| 3. Stock | Tracking units on hand, on order, across warehouses and channels | Counts drift from reality, nobody knows what's at the 3PL versus Amazon, POs live in email | GoodDay | With a forecast you trust, you need the day-to-day system that turns it into orders and keeps the counts honest |
| 4. Cash | Projecting cash from revenue, spend and stock purchases | Doesn't know about payout delays, ad spend, PO timing; built once and never updated | Iris | Needs real books, a real forecast and real PO timing to be worth anything. Last because it depends on the other three |
The mistake most brands make is going in the other direction: buy the fancy planning tool first because that's the one the founder wants, point it at the same messy numbers, and get a very nice-looking dashboard of nonsense. Books first. Always books first. It's the least exciting fix and it's the one that makes the other three work.
1. Books: Finaloop
The reason ecom bookkeeping breaks spreadsheets (and, to be fair, breaks QuickBooks and Xero too) is that nothing arrives in a tidy shape. Shopify pays you out in lumps that don't match the orders. Amazon takes its fees before you see the money and buries them in a 40-column settlement report. Returns land weeks after the sale. Your COGS depends on which batch of stock the unit came from. And none of it arrives labelled. So someone spends the second week of every month reverse-engineering what happened, and by the time the books are right the month is long gone.
Finaloop is bookkeeping built for exactly that mess. It's software plus a team: an AI engine (they call it Rico) that pulls in every channel, bank and payment processor, categorises the lot, reconciles payouts to orders, and tracks inventory and COGS as units move, with a team of accountants behind it checking the work and closing the month. The pitch is real-time books: a P&L, balance sheet and cash flow that are right today rather than right three weeks after month-end. It does accrual or cash, handles B2B and EDI, and plugs into a long list of 3PLs so COGS comes from what actually shipped.
What it costs: published pricing from about $245 a month for brands under $1.5M, stepping up to $995 a month for $6M to $10M, with a 14-day trial. That's the whole bookkeeping function, not a software seat, which is why the reviews keep comparing it favourably to 'QuickBooks plus a bookkeeper plus five integration apps'.
What it's not. It's a closed, opinionated system; you're moving your books into Finaloop, not bolting it onto QuickBooks, and you get less control over the chart of accounts than an accountant might like. It's bookkeeping, not strategy: it'll tell you what your margin was, not what to do about it, and it doesn't do cash forecasting or contribution-margin-by-channel analysis. That's fine. That's not its job. Those are jobs three and four, and they only work because this one's done.
Why first: every tool below is going to pull your revenue, COGS and costs from somewhere. If that somewhere is a spreadsheet someone last tidied in June, the forecast will be wrong with great confidence. Clean books are the foundation. Get this live, get one clean month-end out of it, then move on.
2. Demand: Flagship
The forecasting spreadsheet is usually the one the founder is proudest of and the one that's most wrong. It takes last month's sales, maybe last year's, multiplies by a growth number someone believes in, and spits out a PO. It doesn't know you were out of stock for three weeks in March (so March looks like low demand when it was actually zero supply). It doesn't know the size curve shifted. It doesn't know the promo you're planning. And it has one number per SKU, which means every buy is a single bet with no sense of how confident you should be.
Flagship's whole pitch is literally 'free your inventory from Excel spreadsheets'. It's demand forecasting and merchandise planning built for DTC and apparel-shaped businesses: brands like APL, Mack Weldon, Taylor Stitch and Tracksmith run on it, which tells you the kind of problem it's good at (lots of SKUs, size curves, seasonality, drops). Each product gets its own model, trained on retail and size-curve patterns rather than generic forecasting, and the history is stockout-corrected so a bad supply month doesn't get read as a bad demand month.
The bits that matter for a brand coming off spreadsheets:
- Scenarios, not a single number. Multiple forecast scenarios per product (long-term, short-term, a lighter version) and you pick which one to buy against. That turns 'how many should we order?' into a conversation about risk rather than a guess.
- Size-level buying. Forecasts and POs down to the size, with the variance shown, which is where apparel brands quietly lose the most money.
- Not a black box. You can see the model, the inputs and the reasoning, and override anything. A planner who's been doing this by feel for five years doesn't get replaced; they get a co-pilot that does the arithmetic.
- A PO builder and alerts. The forecast turns into a draft order, and you get told when something's heading for a stockout before it happens. They claim an 80% faster PO workflow versus Excel and a 3 to 9 point drop in stockout rate.
- Your data stays yours. Everything exports to CSV down to SKU level, so if you ever leave, you leave with your history.
It connects to Shopify, Cin7, NetSuite, ShipMonk, Stord, Google Sheets and, usefully for this piece, GoodDay. Onboarding is three to twelve weeks and they don't bill until you're live.
What it's not: it's planning, not operations. It tells you what to buy; it doesn't track what's on the shelf, receive the shipment or allocate stock to wholesale. That's the next job. And it's happiest in brands with real SKU complexity; if you sell six things, you probably don't need it yet.
Why second: once the books are right, you know what actually sold, by channel, net of returns. That's the input a forecast needs. And the forecast is what the next two tools run on: the stock system turns it into orders, and the cash plan turns it into a number you can afford or can't.
3. Stock: GoodDay
The inventory spreadsheet is the one that drifts. It starts accurate, and then a stock take at the 3PL disagrees with it, a wholesale order ships without anyone updating it, a PO arrives short, and six weeks later the number in the sheet and the number on the shelf are two different numbers and nobody's sure which one to trust. Shopify's own inventory tools cover the basics (and with Stocky being shut down, even the basics are getting thinner), but once you've got a 3PL, a wholesale channel and more than one location, you're past what they do.
The traditional answer is an ERP, which for a brand this size means six months of implementation, a consultant, and a system that treats inventory as one module among forty. GoodDay is the other answer: a retail operating system that lives inside the Shopify admin and treats inventory as the atomic unit everything else hangs off. It was built by former operators (the founding team came out of Chubbies) with Shopify engineers, and it's aimed squarely at brands running DTC, wholesale, Shopify POS and marketplaces at the same time.
What it actually does day to day:
- Purchase orders, properly. Create, split, track and receive POs, with vendors, pricing, terms and delivery schedules in one place. The PO Flagship drafts becomes the PO GoodDay raises and receives.
- One ledger, several channels. Virtual inventory segmentation, which means you can ring-fence stock for wholesale, DTC, a retail partner or a pre-order without physically separating it, and stop one channel overselling what another's already promised.
- Counts that stay honest. Real-time reconciliation between Shopify and your WMS or 3PL (it lists 50-plus 3PL and WMS integrations, 100-plus overall), transfers between locations, adjustments, returns, and location-level alerts when something's running low.
- COGS written into Shopify. Automated FIFO cost of goods, pushed into Shopify's own fields, so the margin your books report is the margin the stock actually cost. This is the hand-off back to Finaloop.
- Wholesale and B2B. Live and pre-book sales orders, fulfilment orders, and the EDI and B2B platform integrations wholesale buyers expect.
- Spreadsheet people welcome. GoodDay Sheets keeps Google Sheets as a front end for the ops team that lives there, with the system of record underneath. There's also an MCP connector so you can ask it questions from Claude.
They say most brands are live in weeks, onboarding is done by their team, and support runs through a private Slack channel. They're also offering 90 days free to brands coming off Stocky, which is a reasonable moment to move if you're on it.
What it's not: it won't forecast demand for you (though it now has AI inventory planning features, Flagship goes much deeper), and it's not accounting; it feeds your books rather than keeping them. It's also Shopify-only by design, which is a feature if you're on Shopify and a dealbreaker if you're not.
Why third: you need the forecast before this is worth much, because without one the POs are still guesses, just better-tracked guesses. With the forecast in place, GoodDay is the thing that turns it into stock on shelves and keeps the number in the system equal to the number in the warehouse. Which is, when you think about it, the whole thing the spreadsheet was failing at.
4. Cash: Iris
The cash model is the spreadsheet that gets built once, usually before a fundraise or a scary quarter, and then never quite gets updated. It's got revenue by month, a COGS percentage, a line for marketing and a line for 'other'. It doesn't know that Shopify pays out on a lag, that Amazon holds money for two weeks, that the big PO you just signed off is due in 30 days and the stock won't sell for 90, or that Q4 ad spend front-loads before Q4 revenue arrives. So it says you're fine, and then you're not, and the founder finds out from the bank balance rather than the model.
Iris is FP&A built for brands that sell physical products, which is a different animal from the FP&A tools built for SaaS companies planning headcount. It connects to Shopify, Amazon, your accounting system (QuickBooks, Xero, NetSuite), your 3PL and your ad platforms, pulls it all into one financial model, and keeps it updated daily. The founder built it after being CFO of a CPG brand selling across Shopify, Amazon and retail and not being able to find the tool he needed, which shows in what it prioritises.
What you get that the spreadsheet couldn't do:
- A daily P&L and contribution margin by SKU and by channel. Not 'we made money last month' but 'this product on Amazon is losing money after fees and ads, and this one on Shopify is carrying the business'. They claim the daily P&L lands within half a percent of the closed books.
- Inventory-aware cash forecasting. This is the one that matters for a brand at this stage. The cash plan knows about PO timing, payout delays, lead times and seasonality, so 'can we afford this buy?' gets a real answer.
- Three-statement models and scenarios in plain English. Ask what happens to cash if you push the launch a month, cut ad spend 20% or add a wholesale account, and it builds the scenario rather than you rebuilding the sheet.
- An actual analyst. The managed plans come with a dedicated financial analyst doing variance analysis and weekly check-ins, which for a brand without a finance hire is the difference between having a dashboard and having a finance function.
What it costs: it starts at $2,000 a month and is positioned for brands from roughly $5M up, so it's the most expensive thing in this piece and the one that's most clearly for the upper end of the 'outgrown spreadsheets' range. Under that, a good fractional CFO working off Finaloop's books might be the better buy for now, with Iris as the thing you graduate to.
What it's not: it doesn't keep your books (it sits on top of whatever does), it doesn't manage stock, and it isn't a demand forecaster in the Flagship sense, though it does its own inventory planning for cash purposes. It's the layer that reads everything the other three produce and turns it into a plan.
Why last: a cash forecast is only as good as its inputs. With Finaloop giving it clean revenue and COGS, Flagship giving it a demand forecast it can trust, and GoodDay giving it real PO timing and stock levels, Iris has something worth modelling. Point it at spreadsheets instead and you've bought a very expensive way to be confidently wrong.
How it fits together, and the first 90 days
The nice thing about these four is that they hand off to each other in a loop. Finaloop records what sold and what it cost. Flagship reads that and forecasts what'll sell next. GoodDay turns the forecast into POs, receives the stock, keeps the counts honest and writes the real COGS back into Shopify, where Finaloop picks it up. Iris sits across all of it and tells you whether you can afford the plan. Nothing gets retyped into a spreadsheet at any point, which is the actual goal.
You don't do all four at once. Something like this:
- Month one: books. Move to Finaloop. Get one clean month-end. Resist touching anything else until that's happened; you'll want to, because the books are boring and the forecast is exciting, but the forecast is downstream of the books.
- Month two: demand and stock, together. Start Flagship onboarding and GoodDay onboarding in parallel; they integrate with each other and both take a few weeks. Pick one category to run the first forecast on and one channel to move into GoodDay first. Run the old spreadsheet alongside for one PO cycle so you can see where it was wrong.
- Month three: cash. Once there's a forecast and real PO timing, connect Iris (or, if you're under about $5M, hand the clean books and the forecast to a fractional CFO and revisit Iris next year). Build the first 13-week cash plan from real inputs. Compare it to the old model. Keep whichever one was right, which won't be the old one.
Things to resist. Resist the 'one platform that does everything' pitch; the tools that do all four jobs tend to do all four at a C-minus, and you'll be back in spreadsheets patching the gaps within a year. Resist keeping the old sheets running 'just in case' past the first cycle; two sources of truth is how you got here. Resist buying the cash tool first because it has the nicest dashboard. And resist the ERP consultant who says a brand your size needs NetSuite; some do, most don't, and the six months you'd spend implementing it is six months of decisions made on the old sheets.
The spreadsheets weren't the problem. They were the right answer for a smaller business, and they're the wrong answer for the one you've got now. The fix is four tools doing four jobs, in the right order, talking to each other. Books, then demand, then stock, then cash. Start with the boring one.
A useful tool starts with a clear brief. Explore the profiles below to find the questions worth asking.