Marketing attribution report: a monthly template and what to include

A monthly marketing attribution report is one header line and one table: total revenue that ties to the books, the share matched to a lead, and per channel the leads, matched sales, revenue, share, spend and revenue per dollar, with Direct / Unknown as its own row. Then what is under review, the comparisons and five sentences of notes.

Most monthly marketing reports are activity reports: impressions, clicks, leads, cost per lead and each platform's conversions, one platform per page, ending on a slide that says performance was strong. An owner learns what the marketing did. What it earned they have to work out alone, and usually cannot, because none of those numbers is a sale.

An attribution report has a narrower job: say where this month's revenue came from, and say plainly how much of it nobody can trace. It is shorter than an activity report and harder to make look good, and it is the one that survives an owner with a calculator. Below is the template, with a complete example whose figures add up.

The five parts of a monthly marketing attribution report

A monthly marketing attribution report has five parts, in this order: a header line with total revenue and the match rate, a channel table, a list of matches still under review, a comparison with earlier periods, and three to five notes that explain the changes. Everything else goes in an appendix or nowhere.

1. The header line: total revenue, equal to the books; number of sales; share of revenue matched to a lead; total spend.
2. The channel table: one row per lead source, a Direct / Unknown row and a total.
3. Under review: plausible but uncertain matches, with amounts, not counted until confirmed.
4. Comparisons: last month, the same month last year, and the trailing quarter.
5. Notes: what changed and why, what is too small to judge, and what will be done.

The order is the argument. Coverage comes before channels because every channel figure describes only the part that could be traced. The review list comes before the comparisons because a channel that fell this month may simply have its sales waiting in the queue.

The header line: total revenue first, then the match rate

Open with the total revenue for the month, taken from the same system finance uses, then the share of it matched to a recorded lead. If the total does not equal the figure in the books, fix that before sending anything: a report whose total is disputed will be argued about, not read.

Matched revenue is a subset defined by your own data and rules. Put it second, never in a larger font. "$221,900 of attributed revenue" invites the reader to assume that is the business; "$318,600 of revenue, 69.6% of it matched to a lead" says how much of the business the page describes.

A marketing report is not a securities filing, but the principle transfers: the figure everyone can verify leads, and the figure that depends on your method follows. Which revenue definition and date to use is a conversation to have with finance once; our post on aligning marketing and finance lists where the two usually part. Compute the match rate on revenue, not rows, as our post on match rates explains.

The channel table, column by column

The channel table has one row per lead source, a Direct / Unknown row and a total, and seven columns: channel, leads, matched sales, revenue, share of revenue, spend and revenue per dollar. Each column answers one question, and each has one common way of being misread.

Two rows matter as much as the columns. Direct / Unknown holds every sale whose customer matched no recorded lead: sales, revenue and a share, and no leads or spend. Our post on unattributed revenue explains what it contains and how to shrink it. The total row must equal the header line exactly; if it does not, a sale was dropped or counted twice.

Credit each sale to one channel by one written rule, such as the source of the customer's earliest matched lead, so no sale is counted twice. Keep free channels, such as a Google Business Profile or referrals, in the table with no spend: they are often among the largest rows.

A complete example: September 2026 for a remodeling company

A kitchen and bath remodeler closed 84 sales worth $318,600 in September 2026 and spent $22,500 on four paid channels. 58 of the sales, worth $221,900, matched a recorded lead; 26, worth $96,700, did not. The figures are illustrative; the arithmetic is real and every column adds up.

The header line, as it would open the report: September 2026. Revenue $318,600 from 84 closed sales, agreed with the books. 69.6% of revenue matched to a lead; $96,700 (30.4%) is Direct / Unknown, including 4 sales worth $13,900 under review. Marketing spend $22,500.

The arithmetic. Matched revenue is $79,400 + $36,900 + $31,700 + $28,300 + $23,600 + $22,000 = $221,900, and $221,900 + $96,700 = $318,600. Matched sales are 18 + 11 + 8 + 9 + 5 + 7 = 58, and 58 + 26 = 84. Shares are each row ÷ $318,600: $79,400 ÷ $318,600 = 24.9%, $96,700 ÷ $318,600 = 30.4%, and the column sums to 100.0%. Revenue per dollar is revenue ÷ spend: $79,400 ÷ $9,800 = $8.10, $28,300 ÷ $3,100 = $9.13, $23,600 ÷ $4,200 = $5.62, $22,000 ÷ $5,400 = $4.07.

Three readings come before the ratios. First, the four paid channels brought in $153,300 on $22,500, or $6.81 per dollar; the blended $14.16 divides every sale, referrals included, by spend, and is only a guardrail, as our post on blended ROAS argues. Second, Facebook Ads had the most leads and the fewest sales, a fact about lead quality or follow-up. Third, four of the six channels have fewer than ten sales: one $9,000 kitchen would move direct mail by $2.14 per dollar.

Showing confidence and what is still under review

Show confidence by keeping three groups apart: matches certain enough to count, matches that need a person to confirm them, and sales with no match. Only the first group goes into the channel rows. The second stays in Direct / Unknown and is listed underneath, with amounts and the channel each would join.

Some matches are not safe to count: an email that matches under a different name, a shared office line, a lead from more than a year before the sale. Counting them quietly overstates a channel; dropping them quietly understates it. Listing them is the honest middle.

Then one sentence of consequence: if all four were confirmed, the match rate would be ($221,900 + $13,900) ÷ $318,600 = 74.0%, and Facebook Ads would rise from $4.07 to $28,400 ÷ $5,400 = $5.26 per dollar. That keeps anyone from cutting Facebook on a number that may move next week. For figures that are estimated rather than matched, use words; our post on board reporting suggests three levels, measured, estimated and indicative.

Comparison periods: last month, last year, the trailing quarter

Compare each month with three things: the month before, the same month a year earlier, and the trailing three months against the same three months a year earlier. Last month catches broken tracking. Last year removes most of the season. The trailing quarter gives small channels enough sales to read.

Read the two columns differently. Against August, revenue fell 6.6% ($318,600 ÷ $341,200 = 0.934), which in remodeling is mostly the end of summer. Against September 2025 it rose 10.1% ($318,600 ÷ $289,500 = 1.101) while spend rose 33.9%, a real question about the new money. Our post on seasonal attribution explains why the year-ago column is the one to decide on in a seasonal trade.

Borrow all three habits. Give the month, the year-ago month and the trailing quarter. Mark the figures too small to mean anything. And when a late-closing sale or a confirmed review changes last month's numbers, show the revised figure in a note rather than letting two reports disagree in silence. Where sales close weeks after the lead, mark the latest month as incomplete.

The notes: three to five sentences that explain the changes

The notes say, in plain sentences, what changed and why, what is too small to judge, and what will be done. Each one names a number from the table. Notes that describe activity, such as new creative or launched campaigns, belong in the appendix unless they explain a figure on the page.

Every change has one of four causes, and a note says which: the marketing, the season or the business, the data, or the method. The last two are the ones that get missed, and the ones that cost trust when a reader finds them later. For September:

• Data: the match rate rose 2.8 points on August because web form leads have recorded an email since 18 August. Some of the drop in Direct / Unknown is this fix, not new marketing.
• Under review: Facebook Ads fell to $4.07 per dollar from $5.90; two of its sales, worth $6,400, are under review, and confirmed they would make it $5.26.
• Too small to judge: direct mail had 5 sales from one drop. We will read it on the quarter, after the October drop.
• Season: revenue was 6.6% below August and 10.1% above September 2025; spend was 33.9% higher than a year ago.
• Decision: budgets hold for October. The four sales under review will be confirmed or rejected by 10 October.

If the method changes, such as a new crediting rule or a longer lead export, say so in the first note and show last month on both methods once. A method changed in a bad month and mentioned in a footnote loses a reader for good.

What to leave out of the report

Leave out anything that is not a sale or a cost, anything counted on a different rule from the rest of the table, and anything that makes the table look complete by assumption. That removes most of a typical monthly deck: impressions, click-through rates, platform conversions, and untraced revenue shared out among channels.

• Platform conversions and platform ROAS. Each platform counts on its own rules and windows, so their sum often exceeds total sales. Use them to manage the platform, not here.
• Cost per lead as the headline. It ranks channels by how cheaply they produce inquiries, not sales. In the example, Facebook's leads cost $5,400 ÷ 210 = $25.71 each and Google Ads' $9,800 ÷ 142 = $69.01, and Google Ads brought in almost four times the revenue. Our post on cost per lead covers why.
• Untraced revenue spread across channels. Every row grows by the same factor; nothing new is learned.
• Fractional credit. Splitting a sale 40/40/20 across touches is a model, not a measurement.
• Activity lists. New campaigns and creative tests go in an appendix, unless one explains a number in the notes.

CloseRev builds the channel table in this template from two CSV files you export, closed sales and leads. It matches each sale to leads on normalized phone and email and credits it to the source of the earliest matched lead, written exactly as the lead file's source column has it. Per channel it shows leads, sales, revenue, share of revenue and how many matches need review; only high-confidence matches count automatically, and the rest stay in Direct / Unknown until someone confirms them. Enter monthly spend per channel and it shows revenue per dollar for those channels. The notes are still yours to write.

Presenting it to an owner or a client in five minutes

Present the report in five steps of about a minute each: the total and the match rate, the table from the top, what changed, what is not settled yet, and the decision. Send it the day before, so the meeting is spent on the decision rather than on reading the numbers aloud.

1. Minute one: the total and coverage. "September was $318,600, agreed with the books. We can trace 69.6% of it to a lead; 30.4% we cannot."
2. Minute two: the table, top down. The three largest rows by revenue, their revenue per dollar where they have spend, and the Direct / Unknown row. Nothing else.
3. Minute three: what changed. The notes, starting with any change to the data or the method.
4. Minute four: what is not settled. The review list and the channels with too few sales, and when each will be settled.
5. Minute five: the decision or the ask. Hold, move money, or test, with the amount and the month it will be judged in.

So write for the reader who stops after the header line and the notes, and make sure those two alone tell the truth, untraced share included. Our post on marketing attribution for small business shows the same report at its simplest. For an agency, the hard month is the first, when matched revenue comes in below what the dashboards implied; our post on proving ROI to a skeptical client covers what to say.

A good monthly attribution report is harder to make look good than an activity report, and that is the point. Lead with revenue that ties to the books, say how much of it you can trace, keep the untraced and the unconfirmed visible, compare like with like, and explain every change in a sentence a busy owner will actually read.