When Operating Models Outgrow Spreadsheets

There is a moment in most scaling fintechs where spreadsheets stop being helpful and start being dangerous. Not because Excel is a bad tool, but because it was never designed to carry accountability, judgement, and regulatory pressure at the same time.

This moment rarely announces itself. It arrives quietly, in the gap between what leadership believes is happening and what actually is.

Why Excel hangs on longer than it should

Spreadsheets persist in organisations long past the point where everyone knows they are inadequate. This is not irrational. They solve yesterday's problems remarkably well.

They are fast to build, infinitely familiar, and flexible enough to accommodate almost any new requirement with another tab or column. Most importantly, they create the illusion of control—someone owns the file, everyone can see it, and updates feel immediate.

For early-stage teams moving quickly, this is often exactly right. The spreadsheet adapts as fast as the business does.

The problem emerges later, and it is not where most people expect.

The failure mode is not calculation. It is coordination.

Where it actually breaks

As organisations scale under regulatory scrutiny, spreadsheets begin to fail in predictable ways:

Ownership becomes implicit rather than explicit. Everyone assumes someone else is checking the numbers. Updates happen through conversation rather than process—a Slack message here, an email thread there. Versions multiply. Someone is working off last week's copy while decisions are being made on today's. Email becomes the workflow itself, threading context and approvals through individual inboxes rather than a visible system.

Nobody quite knows what the current state is.

In lightly regulated environments, this friction is tolerable. In fintechs managing KYC workflows, customer onboarding at scale, remediation programmes, or regulatory commitments, ambiguity becomes unacceptable.

The regulator does not accept "we think this is current" as an answer. The board will not tolerate "let me check and get back to you" when asking about portfolio exposure or operational risk.

At this point, the organisation faces a choice it has been deferring.

The operating model gap

Most organisations at this stage do not have a tooling problem. They have an operating model that is no longer being enforced by the way work actually flows.

The model exists—documented in decks, agreed in principle, communicated with intention. But execution happens elsewhere, in the informal layer of spreadsheets, emails, and individual memory.

This is where cadence slips.

Review cycles that were meant to happen weekly become ad hoc. Escalations that should be automatic require someone to notice and raise a hand. Accountability that seemed clear on paper dissolves when it depends on remembering to check a tracker.

The question is not whether people are capable or committed. It is whether the organisation has built an execution layer that makes the operating model real.

What an execution backbone actually does

What is missing at this point is not another system of record. It is a layer that connects decision to delivery, and makes work visible without creating theatre.

The characteristics are straightforward: explicit ownership, so everyone knows who is responsible for what; visible state, so progress and exceptions are legible without asking; traceable updates, so changes and decisions leave a trail; predictable cadence, so reviews and escalations happen when they should, not when someone remembers; and escalation without drama, so exceptions surface naturally rather than through crisis.

This is not about systems. It is about work—how it moves, how it is reviewed, and how an organisation learns from what actually happened rather than what was supposed to happen.

Where platforms fit

In some situations, we use platforms such as Smartsheet as an execution backbone. Not as a replacement for core systems, and not as an automation silver bullet, but as a way of making agreed operating models function day to day.

It is worth being clear about what this is not: not a compliance engine that interprets regulatory requirements, not a substitute for judgement or senior decision-making, and not a replacement for specialist systems like CRMs, ledgers, or payment rails.

What it does is create a stable middle layer—between strategy and systems—where accountability is explicit and cadence is enforced by design rather than discipline.

Why this matters before AI and automation

There is a great deal of energy around AI and automation in financial services at the moment, much of it justified. But most of it is being layered onto processes that are not yet stable.

Without a clear execution model, automation amplifies confusion rather than resolving it. AI has nothing reliable to act on if the underlying work is still happening in email threads and versioned spreadsheets.

Automation needs somewhere sensible to land.

The organisations that will extract real value from these technologies are not necessarily the ones with the most sophisticated tools. They are the ones that can already see, govern, and learn from how work actually happens.

The question worth answering

The question is not which tool to adopt. It is whether the organisation can still see, govern, and learn from how work is actually happening.

When spreadsheets stop being enough, the answer is rarely a big system straight away. It is usually about restoring cadence and accountability first—making the operating model real before adding layers of automation or intelligence on top.

That is a smaller step than most people expect. But it is the one that makes everything else possible.