Measurement

Marketing KPIs that actually help you decide (and the ones you can ignore)

By JCUpdated 8 min read

A marketing KPI is useful if it helps you decide something: where to put the next peso, which campaign to stop, or whether the problem is in marketing or sales. In a service business, six KPIs are usually enough: cost per opportunity, lead‑to‑opportunity rate, time to first contact, close rate, cost per new customer, and sales by channel. Impressions, followers and clicks are good for spotting problems, not for making decisions.

Most reports I see have the opposite problem: many numbers and no decision.

What’s the difference between a metric and a KPI?

A metric is any number you can measure. A KPI is the number that, when it changes, makes you change something.

The test I use is a question: if this number goes up 30% next month, what would you do differently? If the answer is “nothing,” it’s a metric. It can stay on the campaign operator’s board, but not in the report for leadership.

Another clue: a KPI always has an owner. If no one is responsible for moving it, no one will look at it.

Which marketing KPIs help you decide?

These are the ones I include in a service‑business report, in the order I review them:

KPI How it’s calculated Data source Decision it helps you make
Cost per opportunity Channel spend ÷ qualified opportunities that came from that channel Ad platform + CRM Which channel to add or cut budget
Lead‑to‑opportunity rate Opportunities ÷ leads, by channel CRM Which channel brings people who can actually buy
Time to first contact Median minutes between lead entry and first real contact CRM Whether the problem is in follow‑up, not ads
Close rate Sales ÷ opportunities, by salesperson and by channel CRM If you need to work on the proposal, price, or training
Cost per new customer Total marketing spend for the period ÷ new customers Finance + CRM How much you can pay for a customer without losing money
Sales by channel Revenue from customers who came through each channel CRM Which channel to grow next quarter
Sales cycle Average days between first contact and close CRM How long to wait before judging a campaign

Notice the third column. Almost everything comes from the CRM. If your CRM does not record where each lead came from and what stage it is in, you cannot calculate any of these KPIs, even if you have GA4 and Meta set up correctly.

Which metrics can you stop reporting?

I’m not saying you shouldn’t look at them. I’m saying you shouldn’t put them in the budget‑decision report.

Metric Why it doesn’t help decide When to look at it
Impressions and reach They tell you how many people saw something, not if anyone bought When a new campaign generates nothing and you want to know if it’s being shown
Followers They grow with contests and viral content that isn’t related to what you sell Never as a goal; as a sign that your content reaches the right people, check who they are
Reactions and likes They measure likability, not intent To compare which topics resonate more with your audience
CTR (click‑through rate) An ad can get many clicks from people who don’t buy To compare versions of the same ad
Cost per lead It rewards cheap forms even if no one buys To compare within the same channel and audience
Total site traffic It mixes visits from customers with visits from curious people, employees and bots To spot sudden drops
Bounce rate In GA4 it’s the percent of sessions with no interaction; if you don’t track phone clicks, a visit that reads your number and calls counts as a bounce On long pages where you expect reading

About the last: GA4 counts a session as “with interaction” if it lasts more than 10 seconds, has a key event, or has two or more page views (Google definition, reviewed October 17, 2026). Anything that doesn’t meet that counts as a bounce. For a service business, the number that matters is whether that visit ended in a contact, not how many seconds it stayed.

Why cost per lead can be misleading

Because it compares forms, not customers. This example uses made‑up numbers, but the pattern shows up often:

Metric Channel A Channel B
Monthly investment $20,000 $20,000
Leads 100 25
Cost per lead $200 $800
Qualified opportunities 4 8
Cost per opportunity $5,000 $2,500
Sales 1 3
Cost per sale $20,000 $6,667

If you decide based on cost per lead, you give more money to Channel A. If you decide based on cost per opportunity or cost per sale, you give more to B. The difference is threefold.

That’s why I measure all the way to the sale, not just the lead. A lead is a promise. The sale is what pays the payroll.

How many KPIs should a marketing report have?

Between five and seven, on a single page. If you need more, you are probably mixing the decision report with the operational report.

What I do recommend is to separate the sales funnel for each channel:

  1. Lead: someone left their info or wrote.
  2. Opportunity: meets the criteria to buy (budget, need, timeline). Write them down.
  3. Proposal: received a quote.
  4. Sale: signed or paid.

When each step has its number, you can see where the sale falls. If there are many leads and few opportunities, the problem is segmentation or the message. If there are opportunities and few sales, the problem is the proposal, price, or follow‑up. Marketing and sales are one system, and the report must show that.

Where does each data point come from?

Each tool sees part of the picture:

  • Ad platforms know how much you spent and who saw the ad.
  • GA4 knows what people did on your site.
  • The CRM knows who became an opportunity and who bought.
  • Finance knows how much was billed.

They rarely match. Meta can report 40 conversions while your CRM records 12 Meta leads. When that happens, I report both numbers and the difference, then investigate the cause. I don’t pick the one that looks better. If the CRM is well maintained, the recorded sale wins. In how to know which channel brings you customers I explain how to review that difference step by step.

How to build your KPI dashboard in an afternoon

  1. Write your sales funnel and the criteria to move from lead to opportunity. If it isn’t written, each salesperson uses their own.
  2. Check that each lead has a source in the CRM: channel, campaign and, if possible, the ad. Without a source there are no channel KPIs.
  3. Gather the spend per channel for the same period.
  4. Calculate the KPIs for the last 90 days, not just the last month. In services, a September lead may close in November.
  5. Write next to each KPI the decision it triggers. For example: “If a channel’s cost per opportunity doubles the others for two months in a row, cut its budget by 30%.”
  6. Review it once a month with marketing and sales together. If each area reviews its own report, each will explain why the problem belongs to the other.

The goals you set at the start are proposals to validate. Without three months of clean data you don’t know what a good cost per opportunity looks like for your business.

Common mistakes when measuring marketing KPIs

  • Measuring only to the lead. Your ads learn to get cheap forms, not customers.
  • Changing the opportunity definition each month. You can’t compare one month to the next.
  • Mixing sources. Referrals, brokers and ads aren’t measured the same; separate them.
  • Judging in weeks what takes months. If your sales cycle is 60 days, a 15‑day campaign tells you almost nothing.
  • Reporting without a decision. If the results meeting ends without a change in budget, message or process, the report isn’t doing its job.

What I would do this week

  1. Remove from the report everything that doesn’t pass the “what would you do differently?” test.
  2. Check that the CRM records the source of each lead and the date of each stage.
  3. Calculate the cost per opportunity for each channel over the last 90 days and compare it to the cost per lead. If the channel ranking changes, you have found money misallocated.

I implement a CRM and connect ads to sales. If your monthly report has 30 numbers and none tell you what to do, schedule 30 minutes. If you want to start with the technical side, read what to measure in GA4 and Tag Manager.