—
Most marketing teams do not suffer from a lack of data. They suffer from too much of it, arranged in reports that are read, acknowledged and filed without anything changing. A metric earns its place only if a movement in it would lead someone to do something differently: shift budget, rework a message, pause a campaign or fix a page.
Below is a practical way to choose marketing KPIs that support decisions rather than decorate presentations, and to keep the system small enough to be used every week.
Vanity metrics and decision metrics.
A vanity metric is one that looks good when it rises but tells you little about what to do next. Impressions, page views, followers and total email subscribers are the usual examples. They are not worthless, but on their own they rarely connect to revenue, margin or customer retention, and they can grow while the business stands still.
A decision metric has three properties. It is linked to an outcome the business cares about. It can be influenced by actions the team actually controls. And it has an agreed point at which someone will act. Cost per qualified lead, the conversion rate from enquiry to proposal, or the repeat purchase rate within a defined period usually meet this test. The same number can be either kind: website traffic matters little to a firm that wins work through referrals, but a great deal to an online shop.
A simple question helps to sort them: if this number fell by a fifth next month, what would we do? If the honest answer is “nothing in particular”, the metric belongs in a background report, not on the dashboard.
Start from objectives, not from available data.
The most common reason for weak KPIs is that they are chosen from whatever the tools report by default. The better sequence runs the other way: objectives first, then strategy, then tactics, and only then the measures for each level.
01. DEFINE THE BUSINESS OBJECTIVE.
State what the business needs to achieve in a given period, in terms the management would recognise: revenue from a new segment, margin per order, number of new accounts, retention of existing customers. Marketing does not own this objective alone, but it must know which objective it serves.
02. TRANSLATE IT INTO A MARKETING STRATEGY.
Decide how marketing will contribute. Will it attract more of the right prospects, improve conversion of those already interested, raise average order value, or bring existing customers back more often? Each choice implies different measures. A strategy that tries to do everything at once usually ends up measuring everything and improving little.
03. CHOOSE TACTICS AND THEIR MEASURES.
Only now select channels and activities — search advertising, content, email, partnerships, events — and the indicators that show whether each tactic is doing its job. These measures are diagnostic: they explain why the higher-level numbers move, but they are not goals in themselves.
In our own approach, measurement belongs to the control stage. It is designed, however, during analysis and strategy — otherwise one ends up measuring only what happens to be available.
A small KPI hierarchy.
It helps to arrange the chosen metrics in three levels. This keeps each conversation at the right altitude: management looks mainly at the top level, the marketing lead at the middle, and channel specialists at the bottom.
- Business outcome. One to three figures tied directly to the objective, such as revenue from new customers, gross margin or customer retention.
- Marketing outcome. Three to five figures that show marketing’s contribution, for example qualified leads, cost per acquisition, enquiry-to-customer conversion or the share of revenue from returning customers.
- Channel diagnostics. The detailed indicators per channel — click-through rate, cost per click, landing page conversion, email open and click rates — used to find and fix problems.
The rule is that each lower-level metric should have a plausible line of influence to a metric above it. If a channel indicator cannot be linked to any marketing outcome, it is a candidate for removal.
Leading and lagging indicators.
Lagging indicators confirm what has already happened: revenue, new customers, retention over a year. They are the final judgement, but they arrive late, often too late to correct course within a campaign or a quarter. Leading indicators move earlier and signal where the lagging figures are likely to go: the number of qualified enquiries, demo or quote requests, product page visits from target segments, or the share of sales conversations that progress to a proposal.
A useful set contains both. Leading indicators allow the team to react within weeks; lagging indicators keep everyone honest about whether those reactions actually paid off. The relationship between the two should be checked periodically. If qualified leads rise for several months while new business stays flat, either the definition of “qualified” is too loose or the problem lies further down the sales process.
Targets, thresholds and review cadence.
A metric without a target is an observation. A target without a threshold for action is a wish. For each KPI on the dashboard, agree on three things in advance.
01. A TARGET.
Base it on your own history where possible, adjusted for planned changes in budget and activity. External benchmarks can give orientation, but they rarely reflect your pricing, your market and your sales cycle.
02. AN ACTION THRESHOLD.
Define the point at which the owner must investigate and propose a response, for example when cost per qualified lead stays above an agreed level for two consecutive weeks. This turns the dashboard from a report into an early-warning system.
03. A REVIEW RHYTHM.
Match the cadence to the speed of the metric. Channel diagnostics can be checked weekly, marketing outcomes monthly, business outcomes monthly or quarterly. Reviewing slow metrics too often mainly creates noise.
A one-page dashboard.
The dashboard should fit on a single screen or page. If it needs scrolling, it has stopped being a decision tool. A practical layout follows the hierarchy: business outcomes at the top, marketing outcomes in the middle, a compact block of channel diagnostics at the bottom.
For each metric, show the current value, the target, the trend over the last few periods and a clear status: on track, watch, act. Add the name of the owner. A short comment on the past period — two or three sentences on what changed and what will be done — is often more valuable than any additional chart.
Common mistakes.
01. TOO MANY METRICS.
When thirty figures share equal weight, none of them gets attention. Limit the top two levels to around eight metrics in total and keep the rest available for deeper analysis when needed.
02. NO OWNER.
Every KPI needs one person who is responsible for explaining its movement and proposing actions. Shared responsibility for a number usually means that nobody follows it up.
03. NO ACTION THRESHOLDS.
Without agreed limits, every deviation becomes a matter of opinion. Some teams overreact to normal fluctuations; others ignore a decline until it shows in revenue.
04. OPTIMISING THE INDICATOR INSTEAD OF THE OUTCOME.
Once a metric becomes a target, it can be improved in ways that do not help the business: cheaper leads of lower quality, higher click rates from misleading headlines. Pair efficiency metrics with quality metrics so that one cannot be improved at the expense of the other.
In short.
Good marketing KPIs are few, linked to business objectives and owned by someone who knows when to act. Start from the objective, derive a small hierarchy, balance leading and lagging indicators, and agree on targets and thresholds before the numbers arrive. A dashboard that fits on one page and is reviewed regularly will do more for your decisions than any amount of additional data.
Planning your next step? Let’s talk.
0 Comments