Building a KPI Tracking Sheet | The Columns That Matter and Keeping It in Use

You settled on KPIs and built a sheet, but updates stopped within months. Or it does get filled in, yet reading it tells nobody what to do. The cause is usually not the choice of metrics but the columns. This article covers how to build a KPI tracking sheet, and the operating pattern that keeps it alive.

A Sheet That Only Lists Actuals Goes Unused

The typical KPI sheet has three columns: target, actual, attainment.

All that tells you is whether you are behind. Knowing you are behind does not tell you what to do about it.

The result is a standing meeting where numbers get read aloud and nothing follows, until the updates stop too.

The Columns You Need

Six columns is the minimum.

• Metric: what you are watching.
• Target: the figure to reach by period end.
• Actual: what has accumulated so far.
• Projected landing: where this ends up at the current pace.
• Reason for the variance: why target and projection diverge.
• Next action and owner: who does what.

The fourth is the key one.

Why the projected landing column matters

Watching only the actual leaves you acting too late. Sixty percent attainment three quarters in looks fine, but it is a miss if the remaining campaigns do not exist.

With a projected landing column, whether to act now becomes decidable on the spot. The projection does not need to be accurate; what matters is that the gap against target is visible.

Make the variance reason a dropdown

Left as free text, writing it becomes a chore and blanks multiply.

Preparing options keeps the entry going. For example: campaign delay, change in market conditions, wrong assumption, another department's factor, cause unknown.

Keeping "cause unknown" is the trick. Anything marked that way starts with investigation. Force a reason and you get plausible-sounding rationalisations instead.

Monthly or Weekly

What the rows represent depends on the nature of the metric.

The test is how many days pass between acting and the number moving.

• Suits weekly: advertising CPA, site conversion rate. Response shows within days.
• Suits monthly: lead count, opportunity conversion. Weekly views are too noisy.
• Suits quarterly: LTV, branded search volume. These do not move in the short term.

Mix them and a weekly sheet ends up carrying quarterly metrics, so the same figure gets re-keyed every week. That is the classic reason updating becomes a chore.

Separate weekly and monthly sheets.

Five Metrics Maximum

Building a sheet, the temptation is to add metrics. But a sheet with many metrics reliably stops being updated.

Five per sheet is the guide. Beyond that, filling it in becomes the objective and nobody looks at the individual figures.

The test for cutting is: if this number were bad, would we change anything? Numbers that never change behaviour belong on a separate reference sheet.

Keeping the Updates Going

Good columns are not enough without an operating pattern. Three things need settling.

One person updates the numbers

If each owner fills in their own row, one absence leaves the sheet incomplete.

Have one person collect and enter the numbers, and everyone else fill in only the variance reason and the next action. With that split, the numbers are always there.

Set the deadline the day before the meeting

Fill it in during the meeting and there is no time left to discuss the numbers.

Require it by the previous day, and spend the meeting only on the variance reasons and next actions. That turns the standing meeting from a report into a decision.

Decide the procedure for a miss in advance

Starting to think about a response the moment a miss appears makes meetings run long.

Deciding the response by size of miss speeds things up. At 90 percent of target or above, watch it; between 50 and 90 percent, the owner takes a plan away; below 50 percent, discuss changing the campaign on the spot.

Watch-Outs in a Spreadsheet

Starting small in a spreadsheet is fine, but there are traps.

The most common is the file getting duplicated until nobody knows which is current. Fixing a single URL and banning download-and-edit prevents it.

The other is overwriting past periods. Unless rows are added month by month, the trend disappears.

When a Sheet Stops Being Enough

A manual sheet stops working in cases like these.

• Gathering the numbers takes an hour or more each time.
• More people are reading it, and departments have started keeping their own versions.
• Metrics you want to watch weekly have appeared.

The first is the clearest test. Once compiling takes longer than thinking about what the numbers mean, it is time to look at automating it.

Frequently Asked Questions

Can we change the target mid-period?

If you do, keep the original value so the history is visible. Overwrite it and you end up having hit the target by lowering it repeatedly. A single line on why it changed transforms the quality of the review.

Is attainment alone insufficient?

Attainment describes only the past. Only by setting the days remaining alongside where the current pace lands can you judge whether to act now.

People will not fill it in

Check whether what they entered actually gets used in the meeting. Anything filled in but never discussed will be blank next time. Cutting unused columns is the reliable fix.

A metric keeps hitting target

Either the target is low, or the metric is already under control. In the latter case, take it off the sheet and swap in something else. There is little point watching a number that always lands.

Reducing the Time Spent Gathering Numbers

The main reason KPI sheets become a formality is the compiling effort. Pull numbers out of each tool and re-key them every cycle, and eventually nobody does it.

Xtrategy manages campaign schedules alongside budget and KPIs on a single screen. With the numbers arriving automatically, the time goes into the variance reasons and next actions instead.

Summary

• Target, actual, and attainment alone do not lead to action.
• Add three columns: projected landing, reason for the variance, and next action.
• Make the variance reason a dropdown, and keep "cause unknown" as an option.
• Split weekly and monthly sheets by how fast each metric responds.
• Five metrics per sheet at most.
• Concentrate number entry with one person; everyone else writes only reasons and actions.

What helps most is adding the projected landing column. Add one column to the sheet you already use and fill in whether the current pace reaches the target.