Businesses have more financial data than at any point in history.
ERP systems contain invoices, purchase orders and accounting records. Banks provide balances and transactions. Treasury systems contain positions and exposures. CRM systems contain commercial activity. Contracts, procurement systems and operational platforms contain still more information about future financial events.
So, it might seem reasonable to assume that in order to make better financial decisions we simply need to bring all of this data together.
That's certainly useful, but I don't think it's enough, because financial data and Financial Intelligence are not the same thing.
Consider a simple example.
An accounting system contains a customer invoice:
- Amount: £100,000
- Invoice date: 1 September
- Due date: 1 October
- Status: Outstanding
That's financial data. It's accurate, important and useful.
But imagine you're responsible for liquidity. What you really want to know is: When is the £100,000 actually likely to arrive?
The accounting record doesn't necessarily answer that question and it shouldn't be expected to. Its job is to record the invoice correctly.
Understanding what is likely to happen next is a different problem.
From Due Date to Expected Settlement
Suppose we know something else.
This particular customer normally pays between five and eight days after the contractual due date. That changes our understanding.
Perhaps we should expect the cash around 7 October rather than 1 October.
Now suppose we also know that this invoice has been approved by the customer and there are no outstanding queries. Our confidence in receiving it may increase.
Then we discover that the customer's previous three invoices were all paid exactly seven days late. That provides more evidence.
Conversely, perhaps we discover that the invoice has been disputed.
The amount hasn't changed. The due date hasn't changed. The accounting record hasn't changed. But our understanding of the expected cash event has changed substantially.
That's the difference.
Data describes the record.
Intelligence describes what the available evidence means.
Timing and Confidence Are Different Things
There are two concepts here that are easily confused.
When do we expect something to happen?
And:
How confident are we that it will happen?
They're not the same.
We might have a £250,000 receivable that we strongly believe will be paid, but we're uncertain whether it will arrive this week or next.
That's high confidence in the event, but uncertainty around timing.
Alternatively, we might know exactly when a potential £2 million transaction would settle if it proceeds, while remaining uncertain whether the transaction will happen at all.
That's relatively clear timing but lower confidence in the event.
Traditional financial reports struggle to represent this distinction. They tend to give us a value and a date.
Real financial decisions require more context.
A Date Is Not a Prediction
This matters particularly for cash forecasting.
If an ERP says that £10 million of receivables are due next Friday, it's tempting to place £10 million into Friday's forecast.
But the ERP isn't predicting £10 million will arrive on Friday. It's reporting that £10 million is contractually due.
Those are very different statements.
Some customers will pay early, some on time, some late. Some invoices may be disputed. Some payments may settle several invoices together. Others may arrive in instalments.
The due date remains valuable information.
But it is just one piece of evidence about the future, rather than necessarily the future itself.
Evidence Accumulates
Now consider the financial event earlier in its lifecycle.
Perhaps there isn't an invoice yet, but Sales has an opportunity worth £500,000.
That's information about a possible future receipt, but the confidence may be relatively low.
The opportunity progresses and confidence may increase.
A contract is signed. We now have substantially stronger evidence.
Delivery takes place and an invoice is raised. The customer approves it.
Each event tells us something more about the same possible financial outcome.
Eventually the payment reaches the bank. At that point uncertainty disappears.
What was once an expectation has become an observed event.
This suggests that future financial events shouldn't necessarily be treated as isolated records. They can instead be understood as evolving propositions supported by accumulating evidence.
The Same Applies to Expenditure
The principle works just as well in the other direction.
Procurement begins sourcing equipment. A purchase is approved and a purchase order is raised. A supplier accepts the order and goods are delivered.
An invoice arrives. The invoice enters a payment run. The bank confirms settlement.
At each stage the organisation knows something about a future cash requirement.
The amount may become more precise, the expected timing may change, but confidence increases as the event progresses.
Treasury doesn't need to wait until the payment file exists before the expenditure becomes financially significant.
Why Relationships Matter
This also reveals another limitation of looking only at records. Business events have relationships.
- One contract may produce ten invoices.
- One invoice may be settled by three payments.
- One payment may settle several invoices.
- A sales opportunity may result in multiple contracts.
- A purchase order may be amended several times.
If we treat each record independently, we lose part of the financial story.
Financial Intelligence therefore needs to understand not just the records, but how those records relate to one another.
That relationship provides context and context changes meaning.
Start Simple
None of this means an organisation needs to connect every system before obtaining value. Quite the opposite.
A very sensible starting point is Accounts Payable and Accounts Receivable. Bring together what the accounting systems already know and normalise it.
Show Treasury upcoming receipts and payments by date, currency and entity.
That's useful financial visibility immediately.
Then compare those expectations with actual settlement and start understanding payment behaviour.
Then move from contractual due dates towards more realistic expected settlement dates.
Attach confidence to those expectations and explain why an expectation has changed.
Only then, where the value justifies it, begin incorporating earlier evidence from orders, contracts, CRM, procurement and other operational systems.
Financial Intelligence can grow progressively.
Explainability Matters
There is another important requirement.
If a system tells a Treasury Manager: "We expect this £1.2 million receipt on 14 October with 87% confidence" the obvious response should be: "Why?"
A useful Financial Intelligence system needs an answer.
Perhaps:
- the invoice is approved;
- the customer historically pays 6 to 9 days late;
- there are no disputes;
- the last five comparable invoices were paid;
- no contradictory evidence has been detected.
The purpose isn't to create a mysterious score. It's to assemble evidence into an understandable view of the future.
That distinction becomes increasingly important as AI becomes involved.
AI may help discover relationships, interpret documents, identify patterns and assess evidence, but Treasury still needs to understand the basis upon which important financial conclusions have been reached.
From Reporting to Understanding
For decades, we've become increasingly good at collecting and reporting financial data.
That's valuable and isn't going away.
Financial Intelligence asks a different question, not simply: What information does the system contain?
But: Given everything the organisation currently knows, what do we believe is going to happen, when, with what level of confidence, and why?
That's a much more useful question for someone deciding whether to borrow money, invest surplus cash, fund a subsidiary or hedge a currency exposure.
And importantly, the answer isn't static.
As new evidence arrives, expectations change, timing changes and confidence changes.
The organisation's understanding of its financial future should change with them.
That, I think, is the fundamental distinction.
Financial data tells us what we know.
Financial Intelligence helps us understand what that knowledge means.
