Excel is your real route-to-market platform

Yes — for the parts of route-to-market that matter most, Excel is the platform. The official stack captures the clean order-to-cash flow. Everything it can't model — returnable-empties balances, allocation under shortage, trade-promo compliance — lands in a spreadsheet that quietly becomes the system of record. Those files carry no audit trail, no concurrency control, and reconciliation you re-key by hand. The tell is sitting on your weekly commercial call.

Which parts of route-to-market actually run in Excel?

Three domains reliably escape the platform and end up in a workbook. Returnable assets — kegs, bottles, crates, pallets — are a running two-way balance per customer, per SKU, per depot, netted across every delivery and collection and carrying a deposit. Most order-capture systems model a sale, not an asset that comes back; so the empties ledger lives in Excel. Allocation under shortage is a constrained decision with politics attached: when a fast SKU is short, commercial decides who gets seventy percent and who gets starved, by contract, priority and fairness. The platform knows one availability number, not that logic, so the allocation list is a spreadsheet. Trade-promo compliance adjudicates a contract — did the outlet run the display, hit the volume, hold the price — before a rebate pays. The platform books the order; it doesn't judge the promo, so the tracker is Excel too.

The common thread is that these are stateful, exception-heavy and re-cut every quarter — precisely what packaged platforms configure worst and change slowest.

Why does the platform push this work into spreadsheets?

Because Excel is the only part of the estate a commercial analyst can change without a purchase order. Filling the gap properly means a change request, and a change request means the procurement and vendor-roadmap machine — the same dynamic behind the twelve-month purchase of a six-week feature. The money to do it competes with the run-the-business spend that already eats most of the estate, a split we pulled apart in reading your IT budget honestly. And the platform itself moves on a multi-quarter cadence; enterprise implementations run to a median well past a year, as the figures in the 15.5-month median show. None of that helps the empties reconcile this Friday. So someone opens a workbook, because it has a zero-day lead time and nobody has to approve it.

What does the spreadsheet layer actually cost you?

It costs you the properties an event bus would give you for free. There is no record of who changed the allocation, when, or why. Two people edit two copies and you merge them by re-keying. Access is whoever has the file. This is exactly the end-user-computing exposure that SOX auditors flag under their internal-control testing, and the failure modes are well documented. Research collected by the European Spreadsheet Risks Interest Group finds that a large majority of production spreadsheets — commonly cited at around 88 percent — contain errors. Scale bites too: the older .xls format's 65,536-row ceiling caused Public Health England to silently drop nearly 16,000 COVID-19 cases in 2020, and even the modern limit of 1,048,576 rows (as of September 2025, per Microsoft's published specifications) is a hard ceiling a national empties ledger can reach. Manual copy-paste between workbooks is not hypothetical either; it was a documented contributor to a major bank's multi-billion-dollar 2012 trading loss.

How bad is it? Count the spreadsheets on the weekly call.

You don't need an architecture review to size this; you need the attachment list from Monday's commercial call. Count the workbooks that real decisions actually get made from, and rank them by how many people re-key figures out of them. Each one is a feature the platform owes you, and that ranked list is a more honest backlog than any vendor roadmap. Here is the uncomfortable part, and the reason the pattern outlives audits and reorganisations alike: the spreadsheets are usually better at the real job than the platform is. That is why capable people build them. The risk was never that Excel is bad at empties or allocation — it is that Excel is good enough to hide the gap indefinitely, and the gap compounds quietly while it stays hidden.

Where the fix starts

Not with a spreadsheet ban — bans only move the file to a personal drive and delete the paper trail. Treat each surviving workbook as a written specification for a domain object the platform forgot: a returnable-asset ledger, an allocation decision, a promo-compliance record. Model each as first-class state on a shared event bus, with the audit trail, concurrency and access control the file never had. That is the argument behind how we wire every domain onto one event bus.

The brewers who pull ahead won't be the ones with the fewest spreadsheets on any given day — they'll be the ones who treat the spreadsheet count as a live signal. Watch it fall as domains come onto the bus, and watch for the day a new workbook first appears on the commercial call. A fresh shadow spreadsheet is the earliest warning you get that reality has outrun the platform again — and it will, which is exactly why the count, not the roadmap, is the number worth watching.