Budgeting in Excel vs IBM Planning Analytics: What Manual Planning Really Costs Your Company

Excel is free only at first glance. A company pays for its “free” budget with weeks of finance staff time, formula errors and decisions made on yesterday’s numbers. Here is where manual planning loses money, how to work out that cost for your own company, and what IBM Planning Analytics changes.

Excel is not the villain

Let’s be fair first: Excel is an excellent tool. Sketching a model quickly, testing a hypothesis, building a one-person calculation: nothing beats it. The trouble starts when a personal calculator is appointed the corporate planning system.

A budget for thirty departments in Excel is like moving house by bicycle. Possible in theory; in practice you will be riding back and forth all quarter and are bound to lose something on the way.

Where manual planning loses money

1. Time goes into collecting numbers, not analysing them

According to the international FP&A Trends Survey 2024, finance teams spend 45% of their time collecting and validating data and only 35% on insight and decisions. Almost half the working time of expensive specialists goes into stitching spreadsheets together rather than thinking about them.

How a finance department’s working week changes. An illustrative diagram, not survey data.

2. Errors nobody sees

Researcher Raymond Panko spent years gathering the results of audits of real corporate spreadsheets. The finding is sobering: in the more recent audits, errors were found in at least 86% of the files. And it is not about careless staff. People make mistakes in a few percent of complex actions, and a large budget model contains thousands of them.

The best-known example is JPMorgan’s “London Whale” in 2012. A risk model ran on spreadsheets fed by manual copy and paste, and a formula in it turned out to be wrong. The bank lost billions of dollars on those positions. The spreadsheet was not the only cause, but it played its part.

3. “Budget_final_v7_fixed_REALLY.xlsx”

Familiar file name? When the budget lives in email and on a shared drive, nobody knows for certain which version is the latest. A department sends its changes in an old file, the consolidation is built without them, and at the review the CEO sees two numbers for the same line.

For comparison: in IBM Planning Analytics the budget is a single plan with steps, deadlines and owners, not a folder of file versions. Demo data.

4. “What if” scenarios take days

The exchange rate has moved, a supplier has raised prices, the owner asks: “What happens to profit if sales fall by 15%?” In Excel the honest answer is “Give me three days.” In three days the question is either out of date or has been decided by gut feeling.

For comparison: in IBM Planning Analytics budget, actuals and forecast sit in one model, so a “what if” question takes a recalculation, not three days of work. Demo data.

5. It all rests on one person

Every company has the one employee who understands how the master spreadsheet works. While they are around, everything runs. When they are on holiday, ill or leaving, it turns out the company’s budgeting process was stored in a single head.

How to work out the cost for your company

Nobody can give you the exact figure, but you can estimate its order of magnitude in ten minutes. Take three numbers:

1. How many people take part in the budget cycle and the monthly plan-versus-actual review.
2. How many working days a year each of them spends collecting, consolidating and checking spreadsheets.
3. What a day of such a specialist costs, taxes included.

Multiply them. An illustrative example: 20 participants at 25 days a year is 500 person-days, more than two person-years of skilled work spent moving numbers around. And that is only the direct cost. The price of one decision made on a wrong or stale number cannot be written into this formula, and it is usually larger.

What IBM Planning Analytics changes

IBM Planning Analytics is a planning, budgeting and forecasting system built on the TM1 analytical engine. Its job is simple: the company has one model and one version of the numbers.

• One model instead of a hundred files. Departments enter data into a shared system. The consolidation is not “assembled”; it simply exists at any moment.
• Immediate recalculation. Change a rate or a price and the whole model is recalculated. A “what if” answer takes minutes, not days.
• Scenarios side by side. Base, optimistic and stress cases live in one model and can be compared on one screen.
• Approval as a process. You can see who has submitted their budget, who is late, and who changed what. Versions do not get lost.
• AI-assisted forecasting. The system builds a forecast from historical data and explains what drives it. The finance specialist checks and adjusts it instead of drawing a trend by hand.
• Excel stays. People can work with the model from familiar Excel through an add-in. What changes is not the interface but where the data and the logic live.

A dashboard in IBM Planning Analytics Workspace. Demo data.

That last point matters more than it seems. The main fear in a transition is “people won’t want to leave Excel”. They don’t have to: the spreadsheet remains the window, but behind the window there is now a shared model with access rights and a change history rather than somebody’s personal file.

The same model in familiar Excel through the IBM Planning Analytics add-in (Russian interface shown). Demo data.

Seven signs you have outgrown Excel

1. The budget cycle lasts longer than two months.
2. The monthly plan-versus-actual is ready when the next month is already half over.
3. Meetings argue not about decisions but about whose number is right.
4. A “what if” scenario takes several days to calculate.
5. One person understands the master model.
6. The file takes a minute to open and crashes on recalculation.
7. Adding a new department or legal entity is a project in itself.

If you recognise yourself in three or more, the question is no longer “do we need a system” but “how much longer is it worth waiting”.

Where to start

Not by replacing everything at once. The path that works looks like this:

1. Pick one process. Usually the income and expense budget or plan-versus-actual analysis. A pilot on one process shows results faster than “implementing everything”.
2. Put the master data in order. Line items, cost centres, projects. A system will not fix chaos in the structure; it will only speed it up.
3. Connect a source of actuals. A plan without actuals is half the value. The data comes from the accounting system or from a data warehouse. For those already building a warehouse for regulatory requirements this comes almost for free: we wrote about it in the article on ARDFM data marts.
4. Expand step by step. Cash flow, investments, headcount, scenario modelling.

Questions and answers

What is IBM Planning Analytics?

It is IBM’s enterprise planning, budgeting and forecasting system, built on the TM1 analytical engine. It holds a single model of the company and recalculates it whenever the data changes.

Can a budget be run in Excel?

Yes, while the company is small and a few people take part in the process. As departments, legal entities and scenarios multiply, time costs and error risk grow faster than the business itself.

Will employees have to give up Excel?

No. IBM Planning Analytics can be used from Excel through an add-in. The data and calculations are stored in the shared model, not in separate files.

How does IBM Planning Analytics differ from a BI system?

BI shows what has already happened. A planning system lets you enter plans, calculate scenarios and compare plan with actuals. In practice the two are used together.

Which process is best to automate first?

The one where the pain is greatest: most often the income and expense budget or the monthly plan-versus-actual analysis. Other processes are added step by step.

Which companies is IBM Planning Analytics suitable for?

Mid-sized and large companies, holdings and banks: wherever there are many departments, legal entities and participants in the budgeting process, and scenarios are needed.

Sources

• FP&A Trends Survey 2024
• R. Panko. Spreadsheet Errors: What We Know. What We Think We Can Do
• IBM Planning Analytics: IBM product page