KPIs That Don’t Lie: Building Metrics That Reflect Performance

Developing KPIs is one of those tasks that sounds straightforward until you’re sitting in a room with five stakeholders, each with a different idea of what success looks like. Done well, a KPI framework tells you whether your marketing is working. Done badly, it tells you what you want to hear, which is considerably more dangerous.

The core principle is simple: every KPI you commit to should connect, directly or indirectly, to a business outcome. If it doesn’t, it’s a vanity metric dressed up in a spreadsheet.

Why Most KPI Frameworks Fail Before They Start

I’ve sat in more KPI-setting meetings than I care to count, and the failure mode is almost always the same. Someone pulls up a list of metrics from a previous quarter, someone else adds three more from a recent conference they attended, and by the end of the session you have a dashboard with twenty-two metrics, no clear owners, and no agreement on what any of them actually mean for the business.

The problem isn’t the metrics themselves. It’s the process. KPIs are almost always developed backwards: teams start with what’s easy to measure rather than what matters to measure. GA4 shows you sessions, so sessions go on the dashboard. The paid search platform shows you click-through rate, so click-through rate gets reported. These are outputs of the tools you happen to be using, not outputs of strategic thinking.

When I was running iProspect UK, we inherited client reporting that had grown organically over years. Some clients had dashboards with forty-plus metrics, none of which were connected to a coherent story about business performance. The first thing we did in most cases wasn’t to add better data. It was to cut. Getting a client from forty metrics to eight, with clear lines of sight to revenue, was more valuable than any new tracking implementation we could have built.

If you want a broader grounding in how measurement frameworks fit together across the analytics stack, the Marketing Analytics and GA4 hub covers the full landscape, from attribution to data infrastructure.

What Makes a KPI Different From a Metric?

This distinction matters more than people give it credit for. A metric is any quantifiable measure. A KPI is a metric that has been elevated to strategic importance because it indicates progress toward a specific goal. All KPIs are metrics. Not all metrics are KPIs.

The practical difference is that a KPI requires four things a plain metric does not: an owner, a baseline, a target, and a review cadence. Without those four elements, you have a number on a screen, not a performance indicator.

Take organic traffic. It’s a metric. It becomes a KPI only when someone is accountable for it, you know where it started, you’ve agreed where it needs to go, and there’s a scheduled moment when someone will stand up and explain whether it moved in the right direction. Semrush’s breakdown of KPI metrics covers this distinction well if you want a reference framework alongside your own thinking.

The other thing worth understanding is the difference between leading and lagging indicators. Lagging indicators measure outcomes that have already happened: revenue, customer acquisition cost, return on ad spend. They’re important, but by the time they move, it’s often too late to course-correct within the same period. Leading indicators measure activity that predicts future outcomes: pipeline volume, share of search, email list growth rate. A well-designed KPI framework has both.

How to Build a KPI Framework That Holds Up

There’s a sequence to this that most teams skip. They jump straight to choosing metrics without doing the foundational work that makes those metrics meaningful. Here’s the sequence I’ve used consistently across different types of organisations.

Start With the Business Objective, Not the Channel

Every KPI should trace back to a business objective. Not a marketing objective. A business objective. The distinction matters because marketing objectives can be achieved while business objectives are missed. You can hit your traffic target while revenue declines. You can improve your cost per click while customer lifetime value collapses.

The question to ask is: if this metric improves, does the business win? If you can’t answer yes with confidence, the metric belongs in a monitoring report, not a KPI framework.

Early in my career, I worked with a retail client who was obsessed with their email open rate. It was consistently high, and the marketing team reported it proudly every month. When we dug into the data, open rate had almost no correlation with revenue. The audience opening emails were existing customers who rarely bought again. The KPI was technically accurate and strategically useless. We replaced it with revenue attributed to email, segmented by new versus returning customers. The number looked worse immediately, which was exactly the point.

Choose the Right Level of Abstraction

KPIs operate at different levels. Board-level KPIs measure business outcomes: revenue growth, market share, customer acquisition. Marketing-level KPIs measure marketing effectiveness: cost per acquisition, return on marketing investment, brand consideration. Channel-level KPIs measure channel performance: conversion rate, cost per click, organic impressions.

The mistake is mixing levels. When a CEO asks how marketing is performing and the answer is “our click-through rate improved by 12%”, that’s a level mismatch. The CEO is asking a business question. A channel metric is not the answer.

Build your framework in layers. Three to four business-level KPIs at the top. Five to seven marketing-level KPIs in the middle. Channel-level metrics at the bottom, available for operational use but not elevated to the status of KPIs unless they directly drive the layer above.

Set Baselines Before You Set Targets

This sounds obvious and is routinely ignored. Targets set without baselines are guesses. They might be informed guesses, but they’re guesses. Before you commit to any KPI target, you need to know what the current state is, what the historical trend looks like, and what external factors might affect the metric regardless of your activity.

When I joined a business that had been loss-making for several years, the instinct from the board was to set aggressive growth targets immediately. The problem was that nobody had a clear picture of which marketing activities were actually driving revenue versus which were running on historical momentum. We spent the first quarter establishing baselines before we set a single target. It felt slow. It meant we entered the second quarter with targets that were credible rather than aspirational, and we hit them.

Assign Ownership Explicitly

A KPI with no owner is a KPI that won’t be acted on. Ownership means one named person is accountable for the metric, has the authority to influence it, and will be the person who explains performance in review meetings. Not a team. Not a department. A person.

This creates productive tension. When someone knows their name is next to a number that will be discussed in a board meeting, they pay attention to it differently. That’s not a criticism of how people work. It’s just how accountability functions in practice.

Build in Review Cadence From Day One

KPIs need to be reviewed at a frequency that matches the pace at which they can change. Daily review of a metric that moves quarterly creates noise and anxiety. Quarterly review of a metric that should be monitored weekly means you’re always behind. Match cadence to the nature of the metric.

Paid search performance metrics warrant weekly review at minimum. Brand health metrics might warrant monthly. Revenue and customer acquisition cost should be reviewed monthly with a quarterly trend analysis. Build this into your operating rhythm from the start, not as an afterthought.

The Metrics Worth Considering at Each Level

I’m not going to give you a definitive list, because the right KPIs depend entirely on your business model, your stage of growth, and your strategic priorities. What I will do is give you a set of candidates that have consistently proven useful across different types of organisations, with a note on where each one tends to work best.

Revenue from new customers: Separating new customer revenue from total revenue gives you a much clearer picture of whether your acquisition activity is working. Total revenue can grow while new customer acquisition stalls, because existing customers are spending more. That’s not necessarily a problem, but you should know it’s happening.

Customer acquisition cost by channel: Not blended CAC across all channels. CAC by channel, so you can see where acquisition is efficient and where it isn’t. Mailchimp’s guide to marketing metrics covers the mechanics of calculating this if you need a reference point.

Marketing-sourced pipeline: For B2B businesses, the percentage of sales pipeline that originated from marketing activity is one of the most commercially meaningful KPIs you can track. It connects marketing directly to revenue potential in a way that impressions and sessions never will.

Conversion rate by stage: Not a single conversion rate, but conversion rates at each stage of your funnel. Where are people dropping out? Which stages are improving and which are stagnant? This is where operational insight lives.

Return on marketing investment: Calculated properly, not just return on ad spend. ROMI should include all marketing costs, including headcount, agency fees, and technology, against the revenue or margin those activities generated. Semrush’s KPI reporting guide has a useful framework for structuring this calculation.

Brand search volume: The volume of people searching for your brand name is a leading indicator of brand health that most performance marketers undervalue. When I was at lastminute.com, we could see brand search volume move in near real-time following campaign activity. It was one of the earliest signals that something was working or wasn’t.

Common KPI Mistakes That Undermine the Whole Framework

Beyond the foundational errors already covered, there are a handful of specific mistakes I see repeatedly that quietly destroy the value of otherwise well-intentioned KPI frameworks.

Setting targets without considering seasonality. A 20% month-on-month increase in organic traffic sounds impressive until you realise it happened in the month before Christmas in a retail business. Targets need to account for seasonal patterns, or you’ll be celebrating performance that was inevitable and panicking about declines that were equally predictable.

Treating correlation as causation in KPI selection. Two metrics moving together does not mean one is driving the other. I’ve seen businesses track social media follower count as a KPI because it correlated with revenue growth. When they dug into it, both were being driven by a third factor, a major PR campaign, and follower count had no independent effect on revenue at all.

Changing KPIs too frequently. KPIs need time to tell a story. If you change them every quarter in response to whatever the current strategic priority happens to be, you never build the longitudinal data that makes them meaningful. Stability in your core KPI set, with periodic reviews to ensure continued relevance, is more valuable than constant optimisation of the framework itself.

Ignoring data quality. A KPI is only as reliable as the data feeding it. If your tracking is inconsistent, your attribution is broken, or your data pipeline has gaps, your KPIs will reflect those problems without flagging them. Before you commit to any KPI, verify that the underlying data is clean and consistently collected. Understanding how GA4 collects and processes data is a useful starting point if your KPIs are drawing from web analytics.

Using the same KPIs for different audiences. The metrics you report to a CEO are not the same metrics you use to manage a paid search campaign. Using channel-level metrics in board reporting, or using business-level KPIs to brief a media buyer, creates confusion at both ends. Tailor the KPI view to the audience and the decision they need to make.

Connecting KPIs to Your Analytics Infrastructure

A KPI framework is only as good as your ability to measure it consistently. This means your analytics infrastructure needs to be designed around your KPIs, not the other way around. Too many businesses build their measurement stack first and then try to extract KPIs from whatever the tools happen to produce. That’s backwards.

Start with the KPIs you need to track. Then audit whether your current tools can measure them accurately. If they can’t, either fix the tracking or reconsider whether the KPI is measurable in practice. A KPI you can’t measure reliably is worse than no KPI, because it creates false confidence.

For most businesses, the core analytics infrastructure for KPI tracking will include a web analytics platform, a CRM, a paid media reporting layer, and some form of data aggregation tool. GA4 custom event tracking is worth understanding if you’re building KPIs around specific user behaviours on your site, particularly for SaaS or lead generation businesses where standard pageview metrics don’t capture meaningful engagement.

The question of how data flows between these systems, and how consistently it’s collected, is not a technical detail. It’s a strategic one. If your paid media platform reports conversions differently from your CRM, and your CRM reports revenue differently from your finance system, your KPIs will always be contested. Resolving those discrepancies before you lock in your framework saves significant pain later.

Visualisation matters too. A KPI that lives in a spreadsheet that nobody opens is not functioning as a KPI. Whether you use a dedicated dashboard tool or something more integrated, the metrics need to be visible, accessible, and updated at the frequency your review cadence requires. Tools like Tableau can help bring together data from multiple sources into a single view, which becomes increasingly important as your KPI framework matures and spans multiple channels and business units.

If you’re building or reviewing your measurement infrastructure, the full range of analytics considerations, from GA4 configuration to attribution modelling, is covered in the Marketing Analytics and GA4 hub.

When to Review and Revise Your KPI Framework

KPIs are not permanent. They should be reviewed at least annually, and more frequently if the business strategy changes significantly. The test is whether each KPI still measures something that matters to the current strategic direction. If the business has shifted from acquisition to retention, a KPI framework built around new customer acquisition needs to be reweighted.

There are also trigger events that should prompt an immediate review: a major product launch, an entry into a new market, a significant change in competitive landscape, or a structural shift in how the business generates revenue. Waiting for the annual review cycle when the business has fundamentally changed is how you end up optimising for the wrong things for twelve months.

The review process itself should be structured. For each KPI, ask: is this still connected to a business objective? Is the data feeding it still reliable? Has the target been set with current baselines and context? Is the owner still the right person? If the answer to any of those is no, fix it before the next reporting cycle begins.

One final point on this. The best KPI frameworks I’ve seen have one thing in common: they were built by people who were willing to be held accountable for the numbers they chose. When you pick a KPI knowing your name will be next to it in a board presentation, you choose differently. You choose metrics that are honest rather than flattering, measurable rather than aspirational, and connected to the business rather than the channel. That discipline, more than any framework or tool, is what makes KPIs worth having.

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