Monthly Marketing Report Template

Start with who reads the report and what they decide with it: where next month's budget goes, which campaigns continue, whether the quarter's target is still in reach. Pick five to eight numbers that bear on those decisions, and give each a target, usually from your annual marketing plan and budget.

Analytics author Avinash Kaushik defines targets as numbers you set in advance to judge success or failure¹. His example is a YouTube campaign with 1.2 million views. Is that great or awful? Without a target, nobody can say.

Write one definition for each number: the system and field it comes from, the arithmetic, and the similar metric it gets confused with. Sessions are not people. Pipeline is not revenue or a forecast. A cost per lead that divides paid spend by every lead, organic included, is blended, so label it that way.

Then name a system of record for leads and one for pipeline, usually your CRM, so the report already says which tool wins when two disagree. Note the time zone each tool reports in, since a mismatch shifts hours of activity across the month boundary.

Record the settings that change what a number means, and the date of any change. In Google Ads, a new conversion window applies only to conversions going forward², so a month that straddles the change counts conversions two ways. In GA4, changing the reporting attribution model applies to historical and future data³. Last quarter's conversions by channel can shift with no new data at all.

Your newest month is still filling in when you report it. Google says GA4 data processing can take 24-48 hours⁴, and reports may change meanwhile. Google Ads' main conversion columns are calculated based on the time of the click, not the time of the conversion⁵. With a 30-day default window² for Search and Display, a late-March click that converts after your report goes out still adds to March.

Your CRM adds its own lag. A lead's source may be filled in after the record is created, and an opportunity exists only once sales qualifies the lead, which can be weeks later. So a report sent six days after month end sets a six-day-old month beside a 37-day-old one, and the slow numbers look worse than they are.

The fix is to compare like with like: this month against last month as it stood at the same age. If you report on the same day each month, that is simply the figure last month's report published. Put a pull date and a data age (the days between month end and the pull) on every number, so anyone can check that the ages match.

Say which figures are still settling, the way official statistics do. The US Bureau of Labor Statistics marks its latest two months of payroll jobs estimates preliminary and revises each one twice before treating it as final⁶.

Keep every report you publish, since its figures become next month's baseline. Each month, pull the last six months again, compare each figure with what its own report said, and log every change with its cause. Six months of that log shows how long each of your numbers takes to settle.

Give each channel one row: spend, sessions, leads from your system of record, cost per lead where there is spend, and pipeline. Put the pull date and data age on every row. Don't add up the conversions each ad platform reports. Each platform credits its own ads, so one buyer who clicked ads on two platforms can be counted by both.

Show records with no marketing source as a line of their own instead of spreading them across channels. In the template's sample data, 83 of March's 552 new records had no source, so every channel's share is a share of the other 469.

When one channel moves sharply, dig in before the meeting. A Google Ads account audit shows where paid search spend went. Events need care too, since their leads often arrive only when someone uploads the attendee list. A trade show planning checklist that sets a meeting target in advance gives that row a fair benchmark.

Open with any past figure that changed, and its cause. A reader who finds a revised number on their own starts doubting the rest, while the same revision named up front builds trust.

Kaushik says a dashboard for executives needs some data, but what it really needs is insights, actions and business impact⁷. An insight explains why a number moved, not just that it did. Wherever an adjective is doing a number's job, use the number.

End with two or three decisions, each with the figure behind it, an owner and a date. If a decision rests on a number that is still settling, say so. Then cut a one-page summary for leadership from the full report rather than writing it fresh, so the two never disagree.

The template on this page sets this up in five Word documents and five CSV sheets, date columns included. Edit with AI opens it in River, which dates your exports and does the age-matching for you.