Benefits Realisation Tracker: Prove the Savings Are Still There Six Months After Handover

The Benefits Realisation Tracker does one job. It tracks realised benefits against the business case to confirm financial gains are sustained over time. You enter each benefit that your project promised, the planned annual value, the value realised so far this year, the owner and a status. The tool totals the lot and shows you how the portfolio of benefits is travelling.

This sits in the Sustain phase of RDMAICS for a reason. By the time you reach Sustain, the process change is built, the control plan is signed and the project team has moved on. The risk is no longer that the change fails on day one. The risk is that it quietly decays over the following two or three quarters while everyone assumes it is fine.

Use the tracker on a fixed cycle. Monthly works for high value benefits. Quarterly is the minimum for anything you put in a business case. The value of the tool is not any single reading. The value is the series of readings, because a benefit that slips does so gradually and you only see it when you compare one period against the last.

Everything in this guide is built on the sample data that loads with the tool. The case is fictional. It describes an accounts payable improvement project that closed earlier in the financial year and handed four separate financial benefits to the business.

The four benefits are labour cost reduction, late-payment penalty savings, duplicate payment avoidance and early-payment discounts. Three are owned by Finance. One, duplicate payment avoidance, is owned by the AP Manager. Together they were valued at a planned annual total of $235,000 in the business case.

That split matters. A single headline number of $235,000 tells you almost nothing about where risk sits. Four named benefits with named owners tell you exactly who to call when one of them moves. This is the first discipline the tool imposes on you: break the business case into pieces that someone can be accountable for.

The inputs are deliberately few. For each row you provide a benefit name, a planned annual figure, a realised year to date figure, a per cent of plan, an owner and a status. The starred fields are the ones the tool needs to calculate.

In the sample, labour cost reduction carries a planned annual value of $180K with $92K realised year to date and a % of plan of 102%, owned by Finance and marked on track. Late-payment penalty savings is $30K planned against $16K realised at 107% of plan. Duplicate payment avoidance is $15K planned against $8K realised at 107%, owned by the AP Manager. Early-payment discounts is $10K planned against $5K realised at 100%.

Note the difference between the realised year to date column and the % of plan column. Realised year to date is a raw dollar figure taken straight from the ledger. The % of plan is your assessment of how that figure compares against where the benefit should be at this point in the year. They answer different questions and you need both.

Keep the benefit names stable between periods. If you rename labour cost reduction halfway through the year, you lose the ability to compare one reading against the next, and comparison is the whole point.

The tool returns a calculated output panel. For this sample data it shows a planned annual total of $235,000, a realised year to date total of $121,000, and a realised year to date against planned annual figure of 51.5%. It shows 4 of 4 benefits on track and a % of plan range of 100% to 107%.

Read those two percentages as separate signals. The 51.5% is a timing measure. It tells you roughly half the annual planned value has landed so far. Around the middle of a financial year that is what you would expect from benefits that are running to schedule.

The 100% to 107% range is a performance measure. Every benefit sits at or above its own plan for the point in the year. The narrow spread is useful information in itself. When the range widens, one benefit has started behaving differently from the others and that is where you look first.

The tool also offers an AI review that explains the result in plain language. Treat it as a second reading, not as a replacement for looking at the rows yourself.

A benefits tracker that produces no decision is just a report. Here is what you can actually do with the sample reading.

First, you can close the loop with the finance function. The business case promised $235,000. The tracker shows $121,000 realised so far with every benefit at or above plan. That is enough evidence to keep the benefit on the forecast rather than writing it down.

Second, you can choose where to spend your remaining attention. Labour cost reduction is $180K of the $235K planned total. It dominates the portfolio. The other three benefits added together are $55K planned. If you only have time to verify one number properly this quarter, verify the $92K realised against the $180K plan.

Third, you can retire oversight on the small, stable items. Early-payment discounts is $10K planned at 100% of plan. It is doing what it was asked to do and it is small. Move it to an annual check and free up the meeting time.

The sample data shows a healthy portfolio. It is worth being precise about what an unhealthy one looks like, because the warning signs are quiet.

The first sign is divergence in the % of plan range. The sample runs from 100% to 107%, a seven point spread. If one benefit drops out of that band while the others hold, the problem is specific to that benefit and its owner. If the whole band drifts down together, the problem is in the process itself and the control plan is not holding.

The second sign is a realised year to date total that stops moving. The headline 51.5% figure does not tell you whether the benefit is still accruing or whether it stopped accruing two months ago and the total is simply frozen. Only the previous reading tells you that. This is why a single snapshot has limited value.

The third sign is ownership drift. A benefit with an owner who has changed roles, or a benefit that quietly acquires the owner value of the whole finance function rather than a person, is a benefit nobody is actually watching. In the sample, duplicate payment avoidance is owned by the AP Manager. That is a specific role close to the process.

The most common error is entering a forecast as a realised figure. The realised year to date column should contain what the ledger shows, not what you expect the ledger to show by year end. If $92K has landed against a $180K plan, enter $92K. The tool is designed to handle a part-year figure.

The second error is treating the realised against planned percentage as a score. The 51.5% in the sample is not a grade. It is a position in time. Read against the calendar it is unremarkable. Read as a performance figure it looks like failure, and people will react to it as failure if you present it without context.

The third error is marking everything on track by default. Four of four on track is a meaningful statement only if someone was genuinely willing to mark a row otherwise. If status is set by habit, the column carries no information and the tracker stops being a control.

The fourth error is letting the tracker lapse after two readings. Benefits decay slowly. The reading that matters is usually the fifth or sixth one, taken long after the project has been declared a success.

The Benefits Realisation Tracker arithmetic is only as good as the figures you type in. It does not connect to your general ledger. It does not verify that $92K of labour cost reduction actually appeared in anyone's accounts. If the realised figures are estimates dressed up as actuals, the tool will total those estimates and present them with the same confidence as audited numbers.

It also does not establish causation. A realised figure of $16K against a $30K plan for late-payment penalty savings might reflect your improvement. It might reflect a quieter trading period, a supplier contract change or a different team's work. The tracker records the number. It does not tell you why the number is what it is.

It will not warn you on its own. There is no alert when a benefit falls behind. The tool shows you the current state when you run it. Detecting a slip requires you to run it on a schedule and to compare the readings yourself.

Finally, it says nothing about benefits that are not on the list. If the project produced a cost that nobody wrote into the business case, or a benefit nobody thought to record, it sits outside the $235,000 and the tool cannot see it. Use the tracker to hold the recorded benefits to account. Use judgement and a periodic review of the full process for everything else.