Why Supply Chain Must Outgrow Spreadsheets
Spend a few days inside a large retailer, a manufacturer, or a distributor and the real operating system soon reveals itself. The ERP records transactions, dashboards decorate meetings, yet much of the decisive work still happens in spreadsheets. Forecasts are adjusted there, purchase quantities repaired there, transfer plans improvised there, and the last word on a difficult allocation is often buried in a workbook known only to a few practitioners. That arrangement has endured because spreadsheets remain the common working language of planning teams.
I examined the technical side of this question in Chapter 9, and its organizational consequences in Chapters 5, 8, 10, and 11 of Introduction to Supply Chain. The subject is larger than Excel itself. Spreadsheets matter because they capture the prevailing image of supply chain inside most companies: a field of plans, reviews, targets, corrections, and negotiated numbers, prepared for human inspection and repaired by hand whenever reality refuses to cooperate.
Why spreadsheets won
Spreadsheets won honestly. In one place they store data, apply logic, and show results. A practitioner can alter a formula and see the consequence at once. That feedback loop is a serious advantage. It lets supply chain people express local rules without waiting for a release cycle, a budget committee, or a software team that may not understand the business in the first place. For rough analysis, scenario work, and narrow operational fixes, this is excellent engineering disguised as office software.
That is why the spreadsheet-centric camp has always sounded sensible to me. Consultants who defend spreadsheet planning do not merely appeal to habit. They point out, correctly, that Excel is fast to adopt, already present, close to the business, and easier to shape around company-specific rules than many packaged planning suites. Training firms still present spreadsheet fluency as part of the planner’s craft. Even newer enterprise tools preserve much of the same grammar through editable worksheets, scenario comparisons, version history, Power Query imports, and exception-based review. The spreadsheet mind has survived every software cycle because it answers a real need.
I also agree with spreadsheet advocates on a deeper point that is often missed. Supply chain requires ad hoc programming. No static package can absorb the full mess of supplier quirks, rebates, minimum order quantities, routing oddities, calendar distortions, packaging constraints, and commercial exceptions that accumulate in real operations. The persistence of spreadsheets is, among other things, evidence that practitioners need the power to express business logic for themselves.
Where the grid breaks
Trouble begins when a useful local tool becomes the operating model. Workbooks grow by copying. A sheet built for one site is duplicated for another, then patched for a special case, then copied again for a region with slightly different rules. After enough cycles, nobody can say with confidence which formulas drive which decisions. The file still produces numbers, but it stops producing trust. A planning practice that depends on hidden cell references and fragile duplication ages badly.
Supply chain also resists the geometry of the grid. A consequential decision may depend on several suppliers, several transport modes, stepwise rebates, shared capacities, uncertain lead times, partial bills of materials, and priorities that differ by customer or channel. Rows and columns can imitate part of this reality, but only by flattening it, slicing it, and scattering it across helper tabs. Before long the workbook is pretending to be a database, an optimizer, and an application at the same time. None of those roles suit it well.
Once such files become shared working documents, they drift into shadow systems. They contain operational knowledge that the official ledger never captured, yet they lack the discipline expected from anything that affects real money: versioning that can be trusted, permissions that can be enforced, backups that are known to exist, and reproducibility when a bad decision has to be investigated weeks later. Teams end up arguing over the file and the tab instead of arguing over the economics of the choice.
Another confusion follows. Spreadsheet-heavy organizations become attached to planning artifacts: forecasts, safety stocks, service levels, ABC tags, weeks of cover, target fill rates. These figures feel tangible because they compress reality into tidy summaries. Yet they contain no new knowledge beyond the underlying records from which they were calculated. When managers treat those artifacts as the substance of supply chain, they start grooming representations of the problem rather than solving the problem itself.
From there, spreadsheet practice slides naturally into meeting culture. If supply chain is imagined as the art of aligning people around one approved forward view, then the next step is a calendar of reviews devoted to editing and re-editing the same number. Sales and Operations Planning gave this instinct a formal name, but the habit is older than the acronym. It substitutes consensus for judgment. It asks whether the organization is aligned around a plan, not whether scarce inventory, capacity, and cash have been committed where the expected return is highest.
Practitioners often describe spreadsheet planning as giving them control. I understand the appeal. A planner can inspect the workbook, override a suspicious line, and move on. Yet that kind of control belongs to a world of paperwork. It keeps humans close because the machine remains fragile. Systems that emit endless exceptions for people to reconcile are telling us, very plainly, that the machinery cannot be trusted to decide. The spreadsheet then becomes a comfortable refuge, but also a permanent holding pattern.
What replaces the worksheet
What is needed instead is not a larger planning suite with better graphics, nor a campaign to deprive practitioners of programmability. The real task is to keep the expressive power that made spreadsheets useful while moving the logic into an environment where it is explicit, versioned, testable, and fit to run unattended. The workbook diagnosed a real need. It was simply a poor vessel for carrying that need to company scale.
A proper decision system should read authoritative records as they are, preserve uncertainty instead of crushing it into a single forecast, evaluate trade-offs in money, and write back actual decisions: purchase orders, allocations, production starts, price changes. The surrounding discipline is intentionally plain: immutable extracts, reproducible runs, dual-run before cutover, halting when confidence drops, full replay of past states, and audit dossiers that explain why each commitment was issued. This is software built as a productive asset. It gets better when the logic improves, and its improvements compound across every line it touches.
Under that arrangement, people do not disappear. Their work becomes more valuable. The planner no longer spends the day updating formulas and reconciling exceptions. He watches for what data tables still miss: a supplier whose behavior is changing, a regulatory shift that narrows the option set, a commercial decision that alters the value of scarcity, a local disruption that the engine was never meant to infer on its own. The worthwhile intervention is not a handwritten change on a single line. It is a change to the logic, the valuations, or the semantics so that the next thousand decisions improve as well.
My disagreement with the spreadsheet-centric view is therefore sharp, but not casual. Its proponents are right about flexibility, speed, and local knowledge. They are wrong about where those virtues reside. They do not reside in the grid. They reside in the ability to express business logic quickly and to revise it without ceremony. For a narrow problem handled by a small team, a spreadsheet may still be the cheapest way to do that. For a large supply chain, the same habit turns into duplicated logic, weak governance, and an operating model that uses humans as daily patching mechanisms.
Spreadsheets will survive, and they should. They remain useful for rough analysis, ad hoc exploration, and quick sanity checks. But a company that runs its supply chain through spreadsheets has not truly computerized its decisions. It has typed them, circulated them, and asked employees to keep them alive. The next step is to turn decision logic itself into an asset that can be audited, improved, and executed without manual mediation. Once that happens, supply chain begins to look less like administrative housekeeping and more like what it has always been: a disciplined way of committing scarce resources under uncertainty.
