What a Good Monthly Marketing Report Should Tell You

A monthly report exists to answer one question: did the money produce booked work. The numbers that matter in order, the red flags, and how to read one in ten minutes.

A monthly marketing report should tell you what you spent, how many leads that spending produced, how many of those leads were worth having, how many became booked jobs, and what each booked job cost you. Everything else is supporting detail. If you finish reading and still cannot say whether last month’s money produced work, the report has failed, however many charts it contains.

That is a harder standard than most reports are built to meet, which is why so many of them lead with traffic and impressions instead. Here is what a real one looks like, and how to tell it apart from a dashboard screenshot with a cover page.

The one question a report exists to answer

Did the money produce work? That is it. You handed over a budget and a month of your attention, and the report’s job is to close the loop on both. A report that stops at visibility is asking you to accept a chain of assumptions: that impressions became clicks, that clicks became calls, that calls became customers. Every link in that chain is where local marketing usually breaks, and the report is supposed to be the thing that checks them.

This is also the fairest test to apply to your own agency, because it cuts both ways. A month where spend went up and booked jobs did not is worth knowing in the first week of the following month, not in the fourth quarter when you are wondering where the year went. Reporting that connects spend to booked work is how a marketing relationship stays honest in both directions, which is the whole argument behind performance-based marketing.

The numbers that matter, in order

Four numbers carry a local marketing report, and the order matters because each one qualifies the one above it.

Leads by source. Not total leads. Leads split by where they came from: Google Ads, the Business Profile, organic search, Local Services Ads, referral, repeat customer. A single lead total hides the fact that one channel is carrying the month and another is producing nothing.

The qualified share of those leads. Of the calls and forms that arrived, how many were real prospects in your service area for work you actually do? A channel that produces plenty of leads and almost no qualified ones is not performing, it is generating noise, and the raw count will hide that indefinitely.

Booked jobs. How many of those qualified leads turned into work on the calendar. This is the first number in the sequence that has anything to do with revenue.

Cost per booked job. Spend divided by booked jobs, by channel, over a window long enough for the jobs to have closed. This is the number you can compare against what a job is worth to you, and it is the one that tells you where the next dollar should go. It is worth understanding properly, because the arithmetic is easy and the inputs are where people get lost. We wrote about that in detail in cost per booked job.

If a report gives you those four, in that order, split by source, you can run your marketing from it. If it gives you the first one only, you are being asked to guess.

The layer that looks like information and is not

Impressions, clicks, reach, followers, engagement, page views. These are diagnostic tools, and they are genuinely useful when something has gone wrong. If bookings dropped, impressions tell you whether the problem is that fewer people saw you or that the same number saw you and fewer called. That is a real use.

What they are not is a result. Nobody has ever paid an invoice with reach. A report that opens with impressions and buries lead counts on page six has made a choice about what it wants you to look at, and the choice tells you something. The same goes for a follower count on a business where nobody has ever chosen a plumber from a social feed. Metrics that only ever go up, and that no decision depends on, are there to look like progress.

The test is simple. For every chart in the report, ask what you would do differently if that number halved. If the answer is nothing, it should not be leading the report.

One month on its own tells you almost nothing

A single month is a data point, not a trend, and in home services it is a data point taken in weather. Roofing after a storm, HVAC in the first hot week, drain work in the first hard freeze: the month-to-month swing from the season is frequently bigger than anything marketing did. Comparing this July to this June mostly measures the calendar.

A report worth reading shows a rolling view, several months side by side, so you can see direction rather than position. For a seasonal trade it should also put the same month last year next to this one, because that is the only comparison that holds the weather roughly constant. When you have the year-over-year view, a soft month reads correctly as a soft season rather than as a failure, and a strong month gets credited to the season rather than to the agency.

The same logic applies to cost per booked job. It moves with demand. Judging a channel on one expensive month, in the quietest part of the year, is how businesses shut off the thing that was working.

Honest attribution names what it cannot see

Every attribution model is wrong somewhere, and a good report says where. Some of your work has no clean digital trail and never will. The customer who saw your van for six months, searched your name, and called from a saved contact. The one who found you on the map, did not click anything, and dialled. The neighbour referral that happened because a neighbour saw your ad. None of that lands in a channel column cleanly, and a report that assigns all of it anyway is not being more accurate, it is being more confident.

What you want to see is a report that reports direct and branded contacts separately rather than folding them into a paid channel, that says plainly when a jump in “direct” traffic followed a spike in ad spend, and that flags the leads it could not source at all. A line reading “eleven leads with no recorded source, likely map and saved contact” is worth more than a pie chart that adds to a hundred percent by force.

Total honesty here also protects you from the reverse mistake, which is cancelling a channel that was quietly feeding the ones you can see.

Call tracking and form tracking are the plumbing

None of this works without instrumentation, and in home services the phone is where the reporting usually leaks. If calls are not tracked by source, the report cannot tell you which channel produced them, and it will silently under-report whichever channel drives the most calls. That is almost always the Business Profile, which is why profiles so often look weak in reports and strong in reality. The mechanics of setting this up correctly, including the traps specific to the Business Profile, are covered in tracking phone leads.

Forms are easier but not automatic. Every form needs to record where the visitor came from and land somewhere countable, ideally your CRM rather than an inbox. If your agency cannot say how a lead was captured and stored, treat every number downstream of that as an estimate.

Deciding what to capture before any of this is running is its own short exercise, and it is worth doing once: see the one-page measurement plan.

The part most reports skip entirely

What happened, and what changes next month. Numbers describe the past. The report is only worth the meeting if somebody has read them and reached a conclusion.

That means a short piece of writing, in plain language, that says what moved and why, what was tried and whether it worked, what is being changed as a result, and what the change is expected to do. Three or four sentences is enough. “Cost per booked job on the repair campaign rose because two competitors started bidding on emergency terms. We are shifting budget to maintenance queries where the cost is lower, and moving the emergency spend to the hours when we can actually answer the phone.” That is a report doing its job.

Without it, you are being handed the raw material of a decision and left to make it yourself, which is what you paid someone else to do.

Red flags in the report you already receive

The metrics change from month to month. If the headline number is impressions in March, engagement in April and traffic in May, someone is picking whichever line went up. Numbers should be the same every month, especially in the bad months.

There is no commentary. An exported dashboard with a logo on it is not a report. It shows you what happened, not what it means, and interpreting it was the job.

Nothing bad ever appears in it. No month is uniformly good. A report with no losses, no failed test and no cost that went the wrong way is not a picture of your marketing, it is a picture of your marketing’s public relations. Bad news arriving late costs more than bad news arriving.

It reports activity instead of outcome. “We published twelve posts, refreshed four pages and made thirty profile updates.” That is a description of effort. Effort is the agency’s problem, not yours. The only reason to list activity at all is to connect it to something that happened afterwards.

The date ranges do not match. Spend for the calendar month against leads for the last thirty days, or ad platform conversions counted differently from CRM leads, will produce a cost per lead that is quietly wrong. Ask which window each number covers.

How to read one in ten minutes

Skip the front. Find total spend, then total leads, then booked jobs, and write the three on a scrap of paper. Divide spend by booked jobs. Compare that to the same figure from the previous few months rather than to any target in the report.

Then find the split by source and look for the channel that produced leads but no bookings, because that is where money is leaking. Read the commentary next, and check whether it names a change for the coming month or just describes what happened. Last, look for what is missing: unsourced leads, an unexplained gap, a metric that appeared this month and did not exist last month.

If everything reconciles and you understand the plan for next month, you are done. If it does not, the remaining nine minutes are better spent on an email than on the appendix.

The questions worth asking your agency

• How many leads became booked jobs last month, by source?
• What does a booked job cost us on each channel, and how has that moved over six months?
• Which leads could you not attribute, and roughly how many are there?
• What did we try last month that did not work?
• What is changing next month, and what should it do to these numbers?
• Are phone calls tracked by source, and what happens to a lead after it arrives?

An agency that reports this way will answer all six from memory. If the answers require a week and a rebuilt spreadsheet, the reporting is not measuring the work, it is being assembled after the fact to describe it.

If you want a second opinion on the report you are getting now, send it over. It usually takes one read to tell whether the numbers in it are load-bearing.