The essentials
- The myth: watching fifty metrics gives the impression of being in control; in reality attention is diluted, the signals contradict each other and no decision comes out of it.
- The cost: a review of 40 metrics takes an hour and produces fewer decisions than a review of 7 metrics in 15 minutes; the gaps that matter are drowned in the noise.
- The rule: 5 to 8 outcome metrics for the leadership team, 10 to 15 per channel for the team, the rest held in reserve and available on request.
- The rotation: a metric enters the dashboard when a decision depends on it, and leaves it when it has triggered no action for a quarter.
In digital marketing the temptation to track as many metrics as possible is strong: GA4, Google Ads, Meta, Search Console, the emailing tool and the CRM offer hundreds of them, and each one looks useful. Many leaders and marketing managers think a complete dashboard is one that shows everything. The opposite is true: the more metrics you watch, the less you decide. This article explains what the overload costs, how many metrics to keep depending on the role, which ones, and how to rotate the others. The reference metrics are in digital marketing KPIs, and the ones to leave out in vanity KPIs.
What “tracking everything” costs
| Effect | What happens | Consequence |
|---|---|---|
| Diluted attention | Forty charts each get a few seconds | An 8% drop in the conversion rate goes unnoticed between two rises in followers |
| Contradictory signals | Traffic is up, engagement is down, leads are flat, the CPM is falling | Everyone picks the figure that confirms their view; no decision |
| Statistical noise | Out of 40 metrics, 3 or 4 move sharply every month by chance | Actions launched on variations with no cause |
| Cost of production | Each metric calls for a source, a definition, maintenance | The dashboard breaks often, trust falls |
| Long meetings | An hour of reading, fifteen minutes of decision | The leadership team stops turning up |
| Local optimisation | Each metric becomes an objective for someone | The community manager optimises engagement, the traffic manager the CTR, nobody the sales |
How many metrics to keep, depending on the role
| Role | Metrics tracked as a matter of routine | Pace | What stays in reserve |
|---|---|---|---|
| Leadership team | 5 to 8: leads or sales, attributed revenue, cost per customer, channel share, conversion rate, ROI | Monthly | Everything else, on request |
| Marketing manager | 10 to 15: the leadership metrics plus the cost per result by channel, qualified traffic, the reliability of the measurement | Weekly | The channel pages |
| Campaign officer, community manager | 10 to 15 per channel: CTR, CPC, frequency, quality score, engagement by format | Daily to weekly | The platforms' raw reports |
| Agency, analyst | Everything, on request | Depending on the question | Nothing |
These numbers come from a constraint on reading: beyond seven to nine items, a person stops comparing and starts skimming. A metric tracked as a matter of routine is one that is read every time, with its target; a metric in reserve exists in the tool and is looked up when a question calls for it.
Which ones to keep: three tests
- The decision test: which action would change if this metric rose or fell by 20%? If the answer is “none”, it leaves the dashboard. Followers, impressions and page views almost always fail this test.
- The target test: does the metric have a target connected to a business objective? Without a target it has no gap, so there is nothing to read; see adding target objectives to your dashboard.
- The owner test: does someone own it, annotate it, propose actions based on it? A metric without an owner is read by nobody.
A metric that passes the three tests stays; a metric that fails two of them goes. This sort, done once, divides the size of most SME dashboards by three or four without losing a single decision. The link between objectives and metrics is covered in detail in connecting business objectives and marketing KPIs.
Rotating metrics without losing them
- The reserve: the metrics that have been taken out remain available in the tool, on appendix pages or in the platforms; they are not deleted, they are simply no longer read as a matter of routine.
- Entry on demand: a metric enters the dashboard when a decision under way depends on it: the cost per view on YouTube during an awareness campaign, the return rate during a promotional operation.
- Exit at the quarter: a metric that has triggered no action in three months goes back into reserve.
- Diagnostic metrics: CTR, CPM, positions and bounce rate serve to explain a gap, not to detect it; they are looked up when an outcome metric moves out of its range.
- Alerts for the rest: an alert on a metric held in reserve (traffic drop, abnormal spend, broken tag) replaces reading it; it only comes back up if it moves out of its range.
How to run the sort
A two-hour session with the leadership team, the marketing manager and, if needed, the agency. List every metric present in the reports and the dashboard (often 40 to 80). Apply the three tests to each one. Split the result into three groups: leadership routine, team routine, reserve. Write down, for each routine metric, its definition, its target and its owner. Rebuild the dashboard with those metrics alone, the rest on appendix pages. Set the date of the quarterly review at which the list will be revisited. The reading mistakes that survive the sort are covered in the mistakes people make reading marketing metrics, and prioritising actions in prioritising your actions with the dashboard.
How GreenRed helps
Rather than juggling several tools, GreenRed's overview brings these metrics together in a single dashboard, compares them over time and tells you which actions come first. You can try it free, with no card, from the Pricing.
Frequently asked questions
Why shouldn't you track every KPI at once?
Because attention is diluted, the metrics contradict each other, statistical noise triggers actions with no cause, maintenance becomes heavy and meetings drag on without producing a decision. A dashboard of 40 metrics detects fewer significant gaps than a dashboard of 7.
How many KPIs should you track?
Five to eight outcome metrics for the leadership team, read monthly; ten to fifteen for the marketing manager, read weekly; ten to fifteen per channel for campaign officers. The rest sits in reserve in the tools and is looked up when a question calls for it.
How do you choose which KPIs to keep?
With three tests: which decision would change if the metric moved by 20%; does it have a target connected to a business objective; does someone own it and propose actions based on it. A metric that fails two tests leaves the dashboard.
What happens to the metrics that are taken out?
They remain available in the tools and on appendix pages, without being read as a matter of routine. They enter the dashboard when a decision depends on them (awareness campaign, promotional operation) and leave it at the quarter if they have triggered no action. Alerts cover the drifts.
Are diagnostic metrics such as the CTR useless?
No, but they serve to explain a gap, not to detect it. CTR, CPM, positions and bounce rate are looked up when an outcome metric (leads, sales, cost per customer) moves out of its range. Reading them as a matter of routine leads to optimising intermediate metrics at the expense of results.