Every marketing platform reports dozens of numbers, and most of them move without meaning anything.
Impressions rise. Reach fluctuates. Engagement is up this week and down the next. None of it tells you whether to spend more, spend differently, or stop — which is the only reason to look at a number in the first place.
The useful test is the same one that applies to any operational dashboard: what would you do differently if this number changed? If there is no answer, it is not a metric. It is a statistic.
The four numbers that matter
For most businesses, marketing measurement reduces to four things.
Enquiries
The count of people who took the action you want — form submitted, call made, message sent, appointment booked.
This is the number that anchors everything. Traffic without enquiries is a conversion problem. Enquiries without revenue is a qualification problem. Neither is visible without counting this first.
Cost per enquiry
Total spend, including agency fees and your own time where meaningful, divided by enquiries.
This makes channels comparable. It is also the number that reveals when a cheap channel is expensive — a low cost per click producing few enquiries is not cheap.
Enquiry-to-customer rate
What proportion of enquiries become paying customers, ideally split by source.
This is the metric most businesses skip, and it is where the most important discovery usually lives. One channel producing many cheap enquiries that rarely convert can be worse than another producing fewer expensive ones that mostly do. Without this split, you will optimise toward volume and away from revenue.
Value of a customer
What a customer is worth, on a first purchase and over the relationship.
Together with cost per enquiry and conversion rate, this tells you what you can afford to pay for an enquiry — which is the number that makes every budget decision straightforward rather than argumentative.
What is safe to ignore
Impressions and reach. They measure how many times something was displayed, not whether it worked. They rise with spend by definition.
Follower count. Correlates weakly with revenue in most businesses. It is a vanity metric that feels like progress.
Engagement rate, in isolation. Useful as a diagnostic for whether creative is landing. Not useful as a goal, because engagement is not the thing you sell.
Bounce rate, mostly. Frequently misread. A visitor who arrived, got their answer, and left may have been perfectly served.
Rank position as a primary metric. Positions fluctuate constantly and vary by searcher and location. Enquiries are the outcome; rank is one input among several.
Time on page, generally. Longer is not better. Someone finding the answer quickly is a success.
None of these are worthless — several are useful diagnostics when something is wrong. They are poor as targets, because optimising them does not reliably produce customers.
Measure per channel, or you are measuring nothing
An aggregate marketing figure hides everything actionable.
Total enquiries tells you the month was good or bad. Enquiries per channel tells you what to do next month. And enquiry-to-customer rate per channel tells you which of those channels is actually worth its cost.
This requires attribution to exist before the numbers mean anything, which is why measurement precedes optimisation. It is the same argument as what to fix before you spend on ads — without attribution, budget gets allocated by whoever argues most confidently.
Attribution is imperfect, and that is fine
Worth being honest: attribution is never fully accurate. Someone sees a social post, searches your name a week later, clicks an organic result, and calls. Which channel gets credit?
Most models will credit the last click, which understates everything that created awareness. This is a real limitation and not a reason to abandon measurement.
The practical position: use attribution to compare channels directionally rather than to allocate credit precisely. If one channel consistently produces enquiries that convert and another consistently does not, that signal survives the imprecision. Chasing perfect attribution is a project that never finishes and rarely changes a decision.
Simple additions that help more than sophisticated modelling: asking new enquiries how they found you, and comparing periods when a channel was switched off.
Review on a cadence that matches the signal
Different numbers deserve different frequencies, and checking too often is its own failure — daily fluctuation is mostly noise, and reacting to it produces changes that never accumulate.
- Weekly: enquiries, spend, cost per enquiry
- Monthly: enquiry-to-customer rate by channel, organic trend
- Quarterly: customer value, channel mix, what to stop
The most common mistake is reviewing weekly numbers with monthly expectations, then changing course before any change has had time to work.
What we do
Our position is instrument before optimising, which comes from applying engineering discipline to a domain where opinion usually substitutes for data.
Concretely: we establish what an enquiry is worth and whether it can be attributed before recommending any channel or budget. Those two facts determine whether every subsequent decision is evidence or preference — and they are knowable before spending anything on media.
We also do not report on metrics that cannot change a decision. A monthly report full of impressions and reach looks like accountability and provides none.
Frequently asked questions
What if we cannot track phone enquiries?
Ask. "How did you hear about us" recorded consistently is imperfect and far better than nothing. For businesses where phone dominates, this single habit is worth more than most analytics configuration.
How do we measure brand awareness?
Indirectly, and imprecisely. Branded search volume over time is the most practical proxy — if more people are searching your name, awareness is growing. Treat it as a trend, not a precise figure.
Our sales cycle is months. How do we measure anything?
Measure the leading indicators — enquiries and their quality — and accept that revenue attribution lags. Record the source at enquiry so that when a deal closes six months later, you can still trace it.
Should we use a dashboard tool?
Only once you know which numbers matter. A dashboard built before that is a wall of statistics nobody acts on, which is the failure described in operational dashboards people actually use.
How many metrics should we report on?
Four to six. Beyond that, attention disperses and the important movement gets missed among the noise.
Where to start
Take last month. Write down: how many enquiries, from which channels, what they cost, how many became customers, and what a customer is worth.
If you cannot complete that table, the gap is your first priority — not a new channel, and not more spend. Everything else is guesswork until those five figures exist.
Distribution is a system, not a budget covers where measurement sits in the wider picture. If you want help establishing the baseline, talk to us.