A marketing audit checklist an owner can run alone in one afternoon

Say your ad agency reports forty leads last month, and your own enquiry register has eleven. Nobody has asked why, and the invoice for the ads is paid on the fifth. You can find out in one afternoon, alone, by laying five records side by side: your enquiry register, your ad accounts, your website analytics, your CRM or order book, and last year's invoices. The audit is the act of comparing them. Where the numbers disagree, there is something to explain, and the explanation can cost you money.

Before you start: set up the afternoon

Gather the five sources in one place, open a blank sheet, and choose a period. Twelve months is best, because it covers a full season. Write one question at the top: "Where did the orders in this period actually come from, and what did each source cost?" Everything you do for the next three hours answers it.

If a source does not exist, note that and move on. Missing records are a finding, not a failure.

The afternoon in five blocks

Block 1: invoices (40 minutes). Start with money, not marketing. List last year's invoices by customer, with date and amount. Mark which customers are new, which are repeat, and which came by referral. You are looking for where revenue came from, and how concentrated it is. If three customers make up most of the revenue, your marketing question changes.

Block 2: the CRM or order book (30 minutes). For each new customer on the invoice list, find the first record. Note when they first got in touch, what they asked for and who handled it. Note the source if it was written down. If there is no source field, add one now and fill it from memory for the customers you can recall.

Block 3: the enquiry register (40 minutes). Count enquiries by source and by month. Then count how many got a reply the same day, how many got a quote, and how many closed. If you have no register, build one from phone logs and chats for three months. The gap between enquiries and orders is the finding. Our post on how to track lead source in your CRM shows the field to add.

Block 4: ad accounts (30 minutes). Note spend by campaign for the period, and the leads each platform reports. Set those against the register. Platforms count clicks, calls and form fills in their own way, and the register shows what reached you. Where the two disagree, start from the register and ask why.

Block 5: website analytics (30 minutes). Look at where visitors come from, which pages they open, and which actions they take. Compare the busiest month with the month your enquiries peaked. Check that the contact form and the call button actually work from a phone.

The comparison sheet

Fill this in once the five blocks are done:

Divide spend by orders where you can. Where spend is zero, write the time cost if it is material. Do not worry about precision: a rough figure that shows one source costing ten times another is enough to act on.

An invented example: a pump dealer in Kolhapur

This dealer is made up, and so are the proportions. A Kolhapur dealer selling agricultural and domestic pumps spent a fixed amount each month on search ads and a smaller amount on a marketplace listing. The owner believed the ads were the main source.

On the afternoon, the owner found three things. First, the invoices showed that nearly half the revenue came from repeat customers, mostly buyers who rang directly. Second, the register showed that the marketplace had produced a handful of enquiries, all asking for the lowest price, and almost none closed. Third, the ad account reported many "leads", but the register had only a fraction, because most were people tapping the call button for under ten seconds.

The owner did not need a new campaign. The owner stopped the marketplace listing, restricted the ads to the weeks before sowing, and started calling repeat customers before each season. It was not all gain. One contractor who had found the dealer on that marketplace was a real buyer, and the dealer lost that account, though the listing had cost more than the account was worth. The mistake was not the ad spend itself but the absence of a register that could tell the owner whether it worked.

Findings that look bad but are not

Not every gap is a leak. Before you cut anything, check these.

• Ads with few direct orders. A buyer may have seen an ad, then phoned months later. Ask the last five customers whether they remember it before you stop.
• A listing with low-value enquiries. If the enquiries are small but the same customers come back as repeat buyers, the listing may be feeding your base.
• A quiet month on the register. Check the season. Compare the same month a year earlier before you call it a fall.
• A source with no cost. Referrals cost your time and sometimes a gift. Write the cost as "time", not zero.

If you cannot decide, do not cut. Mark it "watch" and put a review date on it.

What to do with what you found

Write three lines at the bottom of the sheet: one thing to stop, one to fix, one to try. Give each a person and a date. Then set a review date and list the numbers you will look at again, taken from the same sheet.

Skip a self-audit if you have no records at all and no time to build them, or if the business is under a few months old and there is nothing to compare. In that case a paid diagnostic or advice from an outsider is more useful than staring at an empty sheet. Our note on sales diagnostic or marketing audit explains how to choose.

If you would like an outsider to look at the same five sources, see marketing audit, and the marketing hub for the planning that follows.

Book the three hours

Put three hours in your calendar this week, gather the five sources the night before, and make the sheet. If you want help acting on what it shows, tell us what you need.