One of the quietest risks I keep seeing in growing finance teams is not the spreadsheet itself.
It is the business logic sitting around the spreadsheet.
The small rules. The exceptions. The manual adjustments. The customer treatment everyone accepts because “that is how we have always done it”. The revenue split that is technically based on the contract, except when sales have agreed something different. The margin report that excludes a warehouse cost because operations says it distorts the trend. The forecast driver that lives in someone’s head because the data is never quite clean enough to trust.
None of this looks dramatic at first.
In fact, it often looks like competence.
A good finance person knows the quirks, fixes the numbers, explains the variance, and gets the report out on time.
But over time, that hidden layer becomes what I call shadow logic.
What Is Shadow Logic?
Shadow logic is the set of business rules that actually drive the numbers, but are not properly documented, governed, automated, or visible to the people relying on the output.
It is not always wrong.
Some of it is commercially sensible. Some of it exists because the source data is messy. Some of it reflects genuine business judgement.
The problem is that nobody can clearly see where the judgement ends and the system begins.
That becomes risky when the numbers are used for board reporting, forecasting, lender updates, pricing decisions, margin analysis, investor conversations, or operational planning.
How Shadow Logic Shows Up in Finance Teams
I saw a familiar version of this recently in a finance process where the headline report looked clean.
The monthly pack had consistent tables. The numbers tied back. The commentary made sense.
But underneath the report, five or six manual decisions were being made every month before the pack was ready.
A product mapping had to be corrected.
A customer segment had to be reclassified.
One cost line had to be moved because the operational system recorded it differently from how finance reported it.
A few new codes had to be checked manually because nobody was sure whether they belonged in the existing hierarchy.
Each decision was reasonable.
Together, they created a process that depended heavily on one person remembering the logic.
That is where CFOs, finance leaders and founders should pay attention.
The Real Risk Is Not Just Human Error
The obvious risk is that someone makes a mistake.
That matters, but it is not the biggest issue.
The bigger risk is that the business starts making decisions from numbers it cannot explain under pressure.
A board asks why margin has moved.
A buyer asks how revenue is segmented.
A lender asks for confidence in the forecast.
A new finance manager joins and cannot tell whether last month’s adjustment was a one-off, a recurring rule, or a correction to bad source data.
When that happens, the business loses time at exactly the moment it needs clarity.
And because the final report looks polished, leadership may not realise how fragile the process underneath has become.
Better Dashboards Do Not Automatically Fix Shadow Logic
The instinct is often to buy a better reporting tool or add another dashboard.
Sometimes that helps.
But a dashboard does not remove shadow logic if the logic is still sitting in email threads, lookup tabs, offline files, manual workbooks, or unwritten judgement calls.
A cleaner visual layer does not fix unclear business rules.
It just makes the uncertainty look nicer.
The first fix is usually more boring and much more useful: write down the rules that matter.
Not every tiny rule. Not a 60-page policy document nobody will read. Just the rules that materially change the numbers or influence a decision.
The Finance Rules Worth Documenting
Start with the rules that affect reporting, forecasting, revenue, margin, customer analysis, cost allocation, and management packs.
For example:
How do we define an active customer?
When does revenue move from pipeline to committed forecast?
Which product hierarchy is the source of truth for margin reporting?
What happens when a transaction has a missing code?
Which manual adjustments are allowed before the monthly pack is issued?
Who approves those adjustments?
Which costs are excluded from operational reporting, and why?
When is a change treated as a correction, and when is it treated as a business rule?
Once those rules are visible, the conversation improves quickly.
You can decide what belongs in the system, what belongs in a controlled model, and what should remain a human judgement with an audit trail.
That is a much better conversation than pretending everything is automated because the final report looks polished.
Spreadsheets Are Not the Enemy
I am not anti-spreadsheet.
Spreadsheets are often the fastest way to understand the shape of a problem. They are flexible, familiar and useful. Most finance transformation projects still start with someone trying to untangle what is really happening in Excel.
The issue is not that spreadsheets exist.
The issue is when the spreadsheet becomes the only place where business rules exist.
At that point, the organisation is building memory in the wrong place.
A practical test I like is this:
If the person who prepares the report was away for two weeks, could another capable finance person reproduce the same result and explain the same logic?
If the answer is no, you do not necessarily have a technology problem yet.
You have a finance logic ownership problem.
How to Start Fixing Shadow Logic
Fixing shadow logic does not need to start with a giant transformation program.
In most businesses, the better starting point is a focused working session.
Pick one important report, forecast, or month-end process. Trace the adjustments. List the judgement calls. Separate data quality fixes from genuine business rules. Then decide which rules should be documented, systemised, approved, or retired.
The process is simple:
Choose one high-value report or forecast.
Identify the manual changes made before it is finalised.
Document the rules and exceptions that drive those changes.
Separate source data problems from business decisions.
Assign ownership for each important rule.
Decide what should be automated, controlled, approved, or removed.
This work is not glamorous, which means it is probably useful. Humanity does occasionally reward the dull but necessary tasks, usually after ignoring them for five years.
But it changes the quality of decision-making quickly.
The Next Automation Opportunity Is Usually Hiding in the Logic
If your finance team is spending too much time explaining numbers after the fact, it may be worth looking for the shadow logic first.
The issue might not be the report.
It might be all the invisible thinking required to make the report behave.
Start with one process. Make the hidden rules visible. You will usually find the next automation opportunity sitting right there.
