On the third Thursday of the month, the sales director presents a revenue forecast that includes a large order for a mining services customer. Two weeks earlier the customer's buyer emailed the account manager to cut the order by forty per cent because of a delay on site. The email reached the account manager's inbox, a note went to operations so that fewer units would be prepared, and the forecast spreadsheet stayed as it was. The gap is only discovered when finance closes the month and the number is short.
This article looks at that operating condition: cancelled or reduced orders that do not update revenue forecasts quickly. It is common where orders are changed by email or phone and where sales, operations and finance each hold their own view of what has been committed.
The problem in plain language
A forecast is only as current as the last change somebody remembered to enter. When an order is amended, several records should change at once: the order itself, the fulfilment plan, the expected invoice and the forecast. In many businesses only the first two change, because they are the ones the customer is waiting on. The forecast sits in a different tool, owned by a different person, and is refreshed on a calendar rather than when something happens.
What the buyer is actually trying to solve
Leadership wants a number it can trust early enough to act on. If a shortfall is visible in week two, the business can pursue substitute orders, adjust purchasing or warn the bank. If it appears in week five, the only choices left are to explain it. The buyer is therefore not looking for a better forecasting model. They want changes to flow into the forecast within a day or two of being agreed, with an audit trail showing what changed and why.
Evidence and system mechanism
A forecast built on stale inputs is a record of the past
The general discipline of sales forecasting depends on the quality and timeliness of the data feeding it. Any method, from a simple weighted list to a statistical model, assumes the inputs reflect the current position. When an order is cancelled in conversation but not in the record, the error is not random noise. It is a systematic overstatement that grows with every change that is not logged.
Revenue recognition depends on what the contract now says
The accounting standard on revenue, IFRS 15, is built around contracts with customers and the performance obligations in them, and it includes requirements for contract modifications. Without going into the technical detail, the principle is relevant for operations: finance cannot treat an amended contract as if it were the original. If the change is not captured as an event the finance team can see, the revenue view is built on a contract that no longer exists. The wider background is covered in the general description of revenue recognition.
Warning signs
- Order changes arrive by email or WhatsApp and are acted on by operations before they are logged anywhere central.
- The forecast is rebuilt manually before each management meeting.
- Finance finds differences between forecast and invoices that sales cannot explain.
- Cancelled orders are discovered when a delivery is refused or an invoice is queried.
Problem owner and why-now trigger
The sales operations lead, or the sales director in a smaller firm, owns the forecast. The operations manager owns the order changes. The financial director owns the consequence: an inaccurate view of cash and profit. Budget for fixing the flow normally sits with the financial director or managing director.
The trigger is often a change in how the business sells. A new pricing model, such as a move from fixed project prices to staged or variable pricing, creates more amendments per order, so the manual routine that coped with a few changes a month breaks when the changes arrive weekly. A missed quarter is the other common trigger.
Economic consequence
The direct cost is decisions taken on wrong numbers. Stock bought for an order that has shrunk ties up cash. Staff scheduled for work that will not arrive cost wages. A forecast that overstates revenue by even a modest margin can lead to a hiring or spending decision the business then has to reverse. Illustratively, if a company with R20 million of monthly forecast revenue carries unrecorded reductions of three per cent, R600,000 of expected revenue does not exist, and nobody knows until the books close. Repeated misses also damage credibility with the board, funders and sometimes customers.
Root cause
- Changes enter through informal channels. Email, calls and chat messages are not structured events and need a person to transcribe them.
- The order has no single record. Sales holds a quote, operations holds a job and finance holds an invoice schedule, as discussed in accepted quotes that do not create fulfilment work.
- Forecast refresh is calendar-driven. It is updated for the meeting, not for the event, which echoes CRM updates that happen only before forecast meetings.
- Nobody owns the amendment. The person who agrees the change with the customer is not the person who updates the forecast.
- Dead or shrunken deals stay at full value. A related pattern is covered in why dead deals stay open in the CRM and distort the sales forecast.
Practical intervention
- Trace three recent changes. Pick three orders that were cut or cancelled and note how long each took to reach the forecast, and who touched it. The delays show where the break is.
- Define an amendment as a recorded event. Every cancellation, reduction or date shift is logged against the order with a reason, an amount and a date, by the person who agreed it with the customer.
- Link the forecast to the order record. The forecast reads order values instead of keeping its own copy. If that is not possible yet, add a daily check that compares the two totals.
- Set a time rule. A change is logged within one working day. Make the rule visible and review misses weekly.
- Separate committed from expected. Show amended orders at their current value, and show the reduction as its own line so that leadership sees movement, not just the new total.
- Report the lag. Track the number of days between agreement and forecast update. Falling lag is the clearest evidence the process is working.
Trade-offs
Requiring a logged reason adds a few minutes to every change, and salespeople may resist when they feel they are doing finance's job. A forecast that moves daily can alarm leadership that is used to a stable monthly number. Tighter links between systems cost money to build, so start with the daily comparison and the one-day rule before investing in integration.
An illustrative example
Illustrative: a Gqeberha packaging manufacturer with 90 staff handled most amendments by email. The sales administrator traced four cut orders and found they took between nine and twenty days to reach the forecast. The company introduced a one-page change form attached to the order, a rule that it be logged the same day, and a Monday comparison between order totals and the forecast. Within two months, the month-end surprise shrank to a few small differences each with an explanation.
Diagnostic questions
- How many days pass between a customer agreeing a reduction and the forecast showing it?
- Where is the first place a cancellation is recorded, and who sees it?
- Can finance tell which part of a forecast change came from amendments?
- What would the last month's forecast have looked like if every change had been logged on the day?
What good looks like
Each change is recorded once, by the person who agreed it, and the forecast reflects it within a day. Leadership sees what moved and why. Month-end differences are small and explained, and decisions about stock, staff and cash are taken on a current number.
Where Mellorca fits
Mellorca's digital systems audit maps how an order change travels today and where it stops. Business automation can then carry amendments from the order record into the forecast so nobody has to retype them.
Method note
The sources describe forecasting practice and the way the revenue standard treats contracts and their modification. They do not measure delay in any particular business. The figures and the example are illustrative rather than a client case, and accounting treatment of a specific contract change should be confirmed with the business's auditors.
Sources and further reading
- Sales operations - Wikipedia (last verified 10 October 2026)
- IFRS - IFRS 15 Revenue from Contracts with Customers (last verified 10 October 2026)
- Revenue recognition - Wikipedia (last verified 10 October 2026)
Related Mellorca reading
- Why account managers cannot see where their customers' orders stand
- Orders are created with values that differ from the accepted quote
- Why revenue leakage only shows up at month-end reconciliation