Marketing metrics that lead to a decision
Choose useful marketing measures: qualified inquiries, conversion rates, customer follow-up, and the limits that make a result meaningful.
A report can be full of numbers and still leave you unsure what to do next. More visits, more impressions, more followers—what changed for the business?
For a small business or startup, a useful measurement plan begins with a decision. You might need to know whether a service page brings relevant inquiries, whether an email helps people find a product, or whether an event promotion reaches people who can attend.
Define the action before the dashboard#
Choose the action that matters to the current project. A service business might track estimate requests. An online store might track completed purchases. A new venture might track conversations with people who fit its intended audience.
Then define the action precisely. A form submission is not automatically a qualified inquiry. An inquiry outside your service area can be useful feedback, but it is different from someone asking for work you can take on.
Follow the path to the outcome#
A simple record can connect the stages:
- How the person heard about you, when known.
- What they viewed or asked about.
- Whether the inquiry matched your service and area.
- Whether you responded and agreed on a next step.
- Whether it became a booking, project, or order.
This can begin in a spreadsheet. The aim is to make patterns visible without building a reporting system nobody maintains. For a local service, record the requested location rather than assuming every website visitor is nearby.
Keep conversion rates in context#
A conversion rate needs a defined numerator and denominator. If you measure inquiries divided by visits, say so. If you measure orders divided by sessions, keep that definition consistent. Always record the period being compared.
Small samples can move sharply. Two inquiries from twenty visits and four from forty have the same rate, but neither alone tells you what will happen next month. Look at the counts alongside the percentages, and note changes to promotions, traffic sources, or the offer.
David’s Nano Bella work includes a reported threefold conversion-rate increase associated with product-page optimization. That is a specific project result from his account, not a forecast for another store. It also does not establish what happened to revenue or the wider business.
A worked example: more inquiries, less useful demand#
Imagine a service business comparing two periods. These numbers are invented to illustrate interpretation, not drawn from an Antilo client or offered as a benchmark. Assume the business uses the same session and inquiry definitions in both periods.
| Measure | Period A | Period B |
|---|---|---|
| Website sessions | 400 | 600 |
| Submitted inquiries | 20 | 36 |
| Inquiries matching service and area | 12 | 12 |
| Submission rate: inquiries ÷ sessions | 5% | 6% |
| Qualified inquiry rate: matching inquiries ÷ sessions | 3% | 2% |
| Share of inquiries that qualify | 60% | 33.3% |
The headline submission rate improved, but the business received the same number of suitable inquiries and more inquiries it could not serve. That does not prove the campaign failed: costs, eventual bookings, and the business’s objectives still matter. It does tell the owner to examine fit before celebrating volume.
Next, inspect the mismatches. Were people outside the service area? Asking for a different service? Responding to an unclear offer? Those answers point to different changes. A broad average cannot tell you which one to make.
Record the time needed for inquiries to become bookings. Comparing a mature group of inquiries with yesterday’s submissions can make the newer period appear worse simply because people have not decided yet.
Define the measure so someone else can reproduce it#
For each measure, write down the event counted, denominator, period, data source, exclusions, and owner. For “qualified inquiry,” specify the service and area criteria rather than letting the definition change with the result. Decide how repeat submissions and obvious spam are handled.
Keep the record proportionate. Aggregate counts and reasons for mismatch are often enough for a marketing review; the report does not need to reproduce private messages or customer contact details. The budget guide shows how to connect these observations to the next scope decision.
Connect local promotion to a useful destination#
If an event handout sends people to a page, give that page a clear purpose. You can use a distinct destination or a consistently named campaign link to help distinguish responses, depending on your measurement setup.
Keep the limits visible: a person may see the handout and later search for your name, or hear about you from several places. “How did you hear about us?” can add context, but it won’t recover every step.
Turn the review into a decision#
Set aside time to ask three questions: what happened, what might explain it, and what should we do next? If inquiries are relevant but responses are slow, the next improvement may be a follow-up process. If visitors reach the right page but seem confused, review the offer and next step.
Choose measures that help you improve the work. Through our digital marketing work , we can help you decide what to measure and how to use what you learn. For a smaller first step, tell us what you’re trying to learn .
