KPIs Are Not a Measurement System. They Are a Management Decision.

A KPI, or key performance indicator, is a quantifiable metric tied to a specific business objective, used to evaluate whether performance is on track. The word “key” is doing the heavy lifting in that definition. Not every metric is a KPI, and treating them as interchangeable is one of the more expensive habits in marketing.

Most businesses have too many KPIs, most of which are not actually key, not actually indicators of performance, and not actually connected to anything a decision-maker can act on. That is not a measurement problem. It is a management problem wearing a measurement costume.

What Makes a Metric a KPI?

The distinction matters more than most people give it credit for. A metric is any quantifiable data point: page views, email opens, social impressions, time on site. A KPI is a metric that has been elevated to strategic significance because it tells you something specific about whether you are achieving a defined objective.

Three things need to be true for a metric to qualify as a KPI. It must be tied to a specific business goal. It must have a target or benchmark against which performance is assessed. And someone must be accountable for it. Remove any one of those three conditions and you have a metric, not a KPI.

I have sat in more quarterly business reviews than I can count where the marketing team presented thirty slides of metrics with no clear line to commercial performance. Impressions were up. Engagement was strong. The click-through rate had improved. And yet revenue was flat. Nobody in the room had the language to challenge it, because everything on the dashboard looked green. That is what happens when you confuse metric collection with performance management.

If you are building or refining your measurement approach, the broader framework around marketing analytics matters as much as the individual metrics you choose. KPIs do not exist in isolation. They are the output of a functioning analytics system, not a substitute for one.

Why Most KPI Frameworks Fail Before They Start

The failure mode I see most often is not that businesses choose the wrong KPIs. It is that they choose KPIs before they have agreed on what success looks like. The metrics come first, and the strategy is reverse-engineered to fit them. That is backwards.

When I was running agencies, one of the first things I would do with a new client was ask them what their three most important business outcomes were for the next twelve months. Not marketing outcomes. Business outcomes. Revenue targets, customer acquisition goals, retention rates, market share objectives. From there, we could work backwards to identify which marketing activities were most likely to drive those outcomes, and which metrics would tell us whether those activities were working.

What I found, consistently, was that the KPIs clients had been tracking bore only a loose relationship to the outcomes they actually cared about. They were tracking what was easy to measure, not what was important to measure. And because the metrics looked respectable, nobody questioned the gap.

HubSpot makes a useful distinction between web analytics and marketing analytics that is relevant here. Marketing analytics connects activity to business outcomes, while web analytics simply reports on what happened on a website. Most businesses are running on web analytics and calling it marketing performance measurement. They are not the same thing.

The Vanity Metric Problem

Vanity metrics are metrics that look good in a report but have no meaningful connection to business performance. Social media follower counts. Raw traffic numbers. Email open rates in isolation. Video views without any downstream conversion data.

The reason vanity metrics persist is not stupidity. It is incentive structure. When marketing teams are measured on metrics they can control and inflate, they optimise for those metrics. If your KPI is impressions, you will get impressions. If your KPI is revenue contribution, you will make different decisions about where to spend your time and budget.

Vanity metrics are not harmless noise. They actively distort resource allocation. I have seen significant budget shifted toward social content programmes because engagement metrics were strong, while paid acquisition channels with clear revenue attribution were underfunded. The engagement numbers felt like evidence of something. They were not. They were activity data dressed up as performance data.

Crazy Egg has a useful breakdown of website KPIs worth tracking that separates engagement signals from conversion signals. The distinction is worth internalising. Engagement tells you that something happened. Conversion tells you that something that mattered happened.

How Many KPIs Should a Marketing Function Have?

The honest answer is fewer than you think. If you have twelve KPIs, you have none. You have a spreadsheet.

A functioning marketing KPI framework typically has three to five primary KPIs at the business level, with supporting metrics beneath each one. The primary KPIs answer the question: is marketing contributing to business performance? The supporting metrics help diagnose why performance is what it is, and where to intervene.

For most marketing functions, the primary KPIs will sit in one of four categories. Revenue contribution, which connects marketing activity to commercial outcomes. Customer acquisition, which measures the volume and cost of new customers. Customer retention or lifetime value, which measures whether you are keeping the customers you acquire. And brand or demand metrics, which measure whether you are building the pipeline that feeds future performance.

That last category is the one most businesses underinvest in measuring, and it is the one that matters most for long-term growth. I spent too many years in the earlier part of my career focused almost entirely on lower-funnel performance metrics: cost per click, conversion rate, return on ad spend. Those metrics are useful. But they predominantly measure your ability to capture existing demand, not your ability to create new demand. A business that only optimises for lower-funnel efficiency will eventually find itself with a shrinking pool of convertible intent and no explanation for why.

Leading vs Lagging KPIs: Why the Distinction Matters

Lagging KPIs measure outcomes that have already occurred. Revenue, customer acquisition numbers, churn rate. They are important because they tell you whether you achieved your objectives. They are limited because by the time you have the data, it is too late to change the inputs that drove the result.

Leading KPIs measure activity or early signals that predict future outcomes. Pipeline volume, lead quality scores, brand search volume, share of voice in key categories. They are harder to define and harder to trust, but they give you something lagging KPIs cannot: time to act.

A well-constructed KPI framework has both. Lagging KPIs tell you what happened. Leading KPIs tell you what is likely to happen. If you are only tracking lagging metrics, you are managing in the rear-view mirror. If you are only tracking leading metrics, you are making assumptions about causality that may not hold.

One of the more useful things I did when I was growing an agency from around twenty people to over a hundred was build a simple leading indicator dashboard alongside the standard commercial reporting. We tracked things like proposal pipeline, client satisfaction signals, and team utilisation rates as forward-looking indicators of revenue performance. It was not perfect, but it gave us a four to six week view of where the business was heading, which meant we could act before problems became results.

Common Marketing KPIs and What They Actually Tell You

It is worth being specific about the most commonly used marketing KPIs, what they measure, and where they are frequently misread.

Cost per acquisition (CPA) measures how much it costs to acquire a customer or conversion. It is useful for efficiency comparisons across channels, but it has a significant blind spot: it says nothing about the quality or lifetime value of what was acquired. A low CPA from a channel that brings in low-value, high-churn customers is worse than a higher CPA from a channel that brings in loyal, high-value customers.

Return on ad spend (ROAS) measures revenue generated per pound or dollar of ad spend. It is widely used and widely misunderstood. ROAS is a ratio, not a profit metric. A ROAS of 4:1 sounds strong until you account for product margins, fulfilment costs, and the portion of that revenue that would have occurred anyway without the advertising. Attribution models that over-credit last-click conversions inflate ROAS figures significantly.

Conversion rate measures the percentage of visitors or leads who complete a desired action. It is genuinely useful as a diagnostic metric, but it is often treated as a primary KPI when it should be a supporting one. Conversion rate tells you how efficiently you are converting the traffic you have. It says nothing about whether you are reaching the right people in the first place.

Customer lifetime value (CLV) measures the total revenue or profit a customer generates over the course of their relationship with a business. It is one of the most strategically important metrics in marketing and one of the least consistently tracked. When I have seen businesses properly integrate CLV into their acquisition strategy, the decisions they make about channel investment and creative approach change substantially.

Brand search volume measures how often people search for your brand name directly. It is an imperfect but useful proxy for brand awareness and demand generation. If your brand search volume is growing, something is working at the top of the funnel. If it is flat while your paid acquisition costs are rising, you have a problem that performance optimisation alone will not fix.

KPIs and Attribution: The Honest Conversation

Attribution is the process of assigning credit for a conversion or outcome to the marketing touchpoints that contributed to it. It is also one of the most contested and frequently misleading areas of marketing measurement.

The problem is not that attribution is impossible. It is that most attribution models are too simple for the complexity of real customer journeys, and the people using them often treat the output as fact rather than as an approximation. Last-click attribution, which assigns all credit to the final touchpoint before conversion, systematically over-credits lower-funnel channels and under-credits the awareness and consideration activity that created the conditions for conversion.

If you are using GA4 for conversion tracking, it is worth understanding how duplicate conversions can distort your KPI data. Moz has a useful piece on avoiding duplicate conversions in GA4 that is worth reading if you are building or auditing your measurement setup. Inflated conversion numbers make KPIs look better than they are, which is the opposite of useful.

I judged the Effie Awards for a period, which gave me an unusual view into how the industry’s most effectiveness-focused campaigns are built and measured. The campaigns that impressed me most were not the ones with the best attribution models. They were the ones where the teams had been honest about what they could and could not measure, had built a coherent theory of how their activity would drive business outcomes, and had tracked against that theory with appropriate humility. Honest approximation beats false precision every time.

For a broader view of how to build a measurement system that connects marketing activity to real business performance, the Marketing Analytics hub covers the full landscape, from GA4 setup to channel-level reporting and attribution frameworks.

Setting KPI Targets: The Benchmark Problem

A KPI without a target is just a metric with ambition. The target is what makes it actionable. And yet target-setting is where most KPI frameworks become politically rather than analytically driven.

The most common failure mode is setting targets based on what feels achievable rather than what is required to meet business objectives. A revenue target of 15% growth requires a specific volume of new customers at a specific average order value. Working backwards from that to marketing KPI targets is straightforward arithmetic. What is less comfortable is when that arithmetic reveals that the current budget or channel mix cannot plausibly deliver the required outcome. Most organisations would rather set a softer target than have that conversation.

Benchmarking against industry averages is useful context but a poor substitute for target-setting. Industry averages mask enormous variance by company size, category, audience, and commercial model. A conversion rate benchmark for a B2B SaaS business tells you almost nothing useful about what your conversion rate should be.

The more useful approach is to benchmark against your own historical performance, adjust for changes in market conditions, and set targets that are connected to a specific business outcome rather than a comparison to peers. If you are setting up measurement infrastructure from scratch, Semrush has practical guidance on how to set up Google Analytics in a way that supports meaningful KPI tracking from the start.

KPIs Across the Funnel: Matching Metrics to Objectives

Different parts of the marketing funnel require different KPIs, and one of the more persistent errors is applying lower-funnel metrics to upper-funnel activity. Measuring a brand awareness campaign by its direct conversion rate is like measuring a fishing net by how many fish it catches on the first cast. It is measuring the wrong thing at the wrong time.

At the awareness stage, relevant KPIs include reach, brand search volume, share of voice, and aided or unaided brand recall where that data is available. These metrics are harder to tie directly to revenue, which is partly why they are undervalued. But without awareness, consideration cannot happen, and without consideration, conversion cannot happen. The funnel is sequential, and the measurement framework should reflect that.

At the consideration stage, relevant KPIs include engagement rate, time on site, pages per session, content consumption depth, and email open and click rates in the context of a nurture sequence. These are diagnostic metrics more than primary KPIs, but they tell you whether your content and messaging is doing the job of moving people from awareness to intent.

At the conversion stage, relevant KPIs include conversion rate, CPA, ROAS, and revenue per visitor. These are the metrics most businesses are most comfortable with, and the ones that receive the most investment in optimisation. They are important. They are also the metrics most likely to be inflated by attribution models that over-credit the final touchpoint.

Moz has a good piece on using GA4 data to shape content strategy that illustrates how consideration-stage metrics can be used to make better decisions about content investment. It is a practical example of connecting mid-funnel KPIs to a specific strategic decision.

How to Build a KPI Framework That Holds Up

The process is less complicated than the industry makes it sound. It requires discipline more than sophistication.

Start with the business objective. Not the marketing objective. The business objective. What does the organisation need to achieve in the next twelve months, and what is marketing’s specific contribution to that? Revenue, customer numbers, retention rates, market share. Get that number agreed and documented before you touch a metric.

Work backwards from the business objective to identify the marketing activities most likely to drive it. This is where strategic thinking matters more than measurement sophistication. If the objective is acquiring a specific number of new customers, which channels, messages, and audiences are most likely to deliver that at the required cost? The answer shapes which metrics matter.

Select three to five primary KPIs that directly connect to the business objective. Assign ownership. Set targets based on what is required, not what is comfortable. Build a reporting cadence that gives you enough frequency to act on the data but not so much frequency that you are reacting to noise.

Beneath each primary KPI, identify two or three supporting metrics that help diagnose performance. These are not KPIs. They are diagnostic tools. If your primary KPI is cost per acquisition and it starts moving in the wrong direction, the supporting metrics help you understand whether the problem is traffic quality, landing page performance, offer relevance, or something else.

Review the framework quarterly. Not to change the KPIs constantly, but to ask whether the metrics you are tracking are still connected to the outcomes that matter. Business priorities shift. Markets change. A KPI that was right six months ago may be measuring something that is no longer strategically relevant.

If you are running conversion tracking through Google Ads, the setup matters more than most people realise. Search Engine Land has a useful historical piece on how Google has approached conversion tracking that provides useful context on how the infrastructure behind your KPI data actually works.

The Accountability Question Nobody Asks

KPIs without accountability are aspirations. The measurement framework is only as useful as the decisions it informs and the people who are responsible for acting on it.

One of the more uncomfortable things I have observed across twenty years of working with marketing teams is how rarely KPI ownership is genuinely assigned. Metrics are tracked collectively, reported collectively, and when performance falls short, accountability diffuses across the team in a way that makes it impossible to identify what needs to change. Everyone is responsible for everything, which means nobody is responsible for anything.

Effective KPI frameworks assign specific ownership to specific individuals. Not teams. Individuals. The person who owns the KPI is responsible for understanding what is driving performance, communicating risks early, and proposing interventions when targets are at risk. That is a different job than pulling a report and presenting it in a meeting.

If you are working with data from multiple channels and need to bring it together in a coherent view, tools like Sprout Social’s Tableau integration illustrate how cross-channel data can be consolidated into a single reporting view. The technology is the easy part. The accountability structure is the hard part, and it is the part that determines whether the data actually changes anything.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions