KPIs Are Lying to You. Here’s Why.

A KPI, or Key Performance Indicator, is a measurable value that shows how effectively a business or marketing activity is progressing toward a defined objective. The operative word is “key.” Not every metric is a KPI. A KPI is the specific number that tells you whether you are winning or losing against a goal that actually matters to the business.

In practice, most organisations treat the definition loosely. They track dozens of metrics, call all of them KPIs, and end up with dashboards that generate a lot of activity and very little clarity. The problem is not the data. The problem is the selection.

I spent a good portion of my career inside agencies where KPI conversations happened constantly. Client reviews, quarterly business reports, new business pitches. The language of KPIs was everywhere. What was rarer was the honest question: does this number actually tell us whether the marketing worked? Not whether it ran, not whether it was delivered on time and on budget, but whether it moved the business forward in a way that justified the investment.

What Makes a Metric a KPI?

The distinction matters more than most people give it credit for. A metric is any quantifiable measurement. A KPI is a metric that has been deliberately selected because it tracks progress toward a specific, significant goal. That selection process is where the thinking happens, and where most organisations cut corners.

Consider website traffic. Traffic is a metric. It becomes a KPI only when there is a clear, documented reason why increasing traffic will lead to a business outcome you care about. If your conversion rate is stable and your revenue grows proportionally with traffic, then traffic is a reasonable KPI. If your traffic is growing while revenue is flat, traffic is not a KPI. It is a vanity metric wearing a KPI badge.

The test I have always applied is simple: if this number improves but nothing else changes, does anyone actually benefit? If the answer is no, it is not a KPI. It is a comfort metric, something that makes reporting feel productive without requiring hard thinking about whether the work is delivering value.

If you are building or reviewing your measurement framework, the broader context of marketing analytics is worth understanding before you start assigning KPIs to channels and campaigns. The framework shapes the quality of every individual indicator within it.

Why Most KPI Frameworks Are Broken

When I took over a loss-making agency and started working through the reporting infrastructure, one of the first things I found was a client dashboard with 47 metrics on it. Forty-seven. The client had been receiving this report monthly for two years. When I asked the account team which of those numbers had ever driven a decision, they could name three. The other 44 existed because no one had ever been brave enough to remove them.

This is not unusual. It is the norm. KPI frameworks grow by addition. Someone asks for a new metric to be tracked. It gets added to the dashboard. It is never removed. Over time the dashboard becomes a museum of past concerns rather than a live instrument for current decisions.

The consequence is that attention gets diluted. When everything is a KPI, nothing is. Teams spend time explaining movements in numbers that no one will act on, and the metrics that genuinely indicate business health get lost in the noise.

There is also a structural incentive problem. Agencies and internal teams alike tend to surface KPIs that make their work look good. Click-through rates, impressions, engagement rates. These are easy to improve and hard to connect directly to revenue. They survive in reporting packs because they fill space and rarely generate uncomfortable questions. The best website KPIs are the ones that create accountability, not the ones that generate applause.

The Attribution Problem Hiding Inside Every KPI

Here is the issue that took me longer than it should have to fully appreciate. A KPI is only as honest as the attribution model sitting behind it. If you are measuring conversions using last-click attribution, you are not measuring the impact of your marketing. You are measuring the last touchpoint before a purchase, which is a very different thing.

Earlier in my career I was a true believer in lower-funnel performance metrics. Cost per acquisition, return on ad spend, conversion rate. These numbers felt concrete. They had a directness that brand metrics lacked. The problem is that I was measuring the capture of demand that already existed, not the creation of new demand. Someone who types a brand name into Google and clicks a paid search ad was probably going to buy anyway. Crediting that conversion to the paid search campaign is not wrong exactly, but it is not the full picture either.

Think about a clothes shop. The moment someone tries something on, the probability of them buying increases dramatically. But the sale does not get credited to the changing room. It gets credited to the till. If you only measured till transactions, you would never invest in good changing rooms. That is roughly what last-click attribution does to upper-funnel marketing.

Google has made conversion tracking progressively more accessible over the years, which is genuinely useful. But more accessible data does not mean more honest data. Easier conversion tracking still requires careful thinking about what a conversion actually represents and which channel deserves credit for it.

How to Choose the Right KPIs for Marketing

The selection process should start with the business objective, not with the available data. Most teams do it the other way around. They look at what their tools can measure and work backward to find a business rationale. That produces metrics that are convenient rather than meaningful.

A more defensible approach works in three stages.

Stage One: Define the Business Objective

Not the marketing objective. The business objective. Revenue growth, market share, customer retention, new customer acquisition. Something the CFO and CEO would recognise as meaningful without needing it explained. Marketing exists to serve business goals. KPIs that cannot be connected to those goals are not KPIs.

Stage Two: Identify the Marketing Levers

What marketing activity, if it performs well, will move the business objective? This is where you map the relationship between marketing outputs and business outcomes. Reach more people in your target audience, conversion rate improves, revenue grows. That chain needs to be explicit and defensible, not assumed.

Stage Three: Select Metrics That Track the Levers

Now you can look at what your tools can measure. The question is not “what can we track?” but “what metric will tell us whether the lever is working?” For email marketing, that might be click-to-conversion rate rather than open rate. Open rate tells you about subject lines. Click-to-conversion tells you whether the email is driving behaviour that matters. Email marketing reporting that focuses on the former at the expense of the latter will consistently overstate performance.

Keep the list short. Three to five KPIs per channel is a ceiling, not a floor. If you cannot articulate in one sentence why each one matters to the business, cut it.

Leading vs Lagging KPIs

This distinction is underused in marketing. A lagging KPI measures an outcome that has already occurred. Revenue, customer acquisition cost, return on ad spend. These are important but they tell you what happened, not what is about to happen. By the time a lagging KPI signals a problem, you are already behind.

A leading KPI measures something that predicts a future outcome. New qualified leads entering the pipeline. Share of search within a category. Branded search volume. These are harder to define and harder to defend in a boardroom, but they give you time to act.

The best measurement frameworks include both. Lagging KPIs confirm whether the strategy worked. Leading KPIs tell you whether it is working. Running a business on lagging indicators alone is like driving by looking in the rear-view mirror.

When I was growing an agency from around 20 people to over 100, the lagging indicators were obvious: revenue, margin, headcount utilisation. The leading indicators were harder to define but more valuable: proposal win rate, client satisfaction scores, pitch volume. If the leading indicators started moving in the wrong direction, I had six to twelve months to respond before the lagging indicators caught up. That time is what separates proactive management from crisis management.

KPIs Across Different Marketing Channels

Different channels serve different purposes in the customer experience, and their KPIs should reflect that. One of the most persistent measurement errors I have seen is applying the same performance standards across channels that are doing fundamentally different jobs.

Paid search captures intent. Someone is actively looking for what you sell. The appropriate KPI is conversion-focused: cost per acquisition, return on ad spend, conversion rate. These are reasonable because the channel is operating at the bottom of the funnel where purchase intent is already established.

Display advertising builds awareness. It reaches people who are not yet in-market. Measuring it on last-click conversions is like judging a billboard by how many people immediately walked into a shop after seeing it. The right KPIs for awareness channels are reach, frequency, brand search uplift, and over longer timeframes, market share movement.

Content marketing operates across the full funnel. A piece of content might drive awareness, educate a prospect, and support a conversion decision across a period of weeks. Using GA4 data to inform content strategy can help identify which content is genuinely contributing to the path to purchase rather than just generating traffic that goes nowhere.

Email sits closer to retention and conversion. The metrics that matter here are engagement with intent signals: clicks on product pages, repeat purchase rate among email-engaged customers, reactivation rate for lapsed buyers. Standard marketing metrics give you a baseline, but the real value comes from segmenting performance by audience behaviour rather than treating all subscribers as equivalent.

The Vanity Metric Problem

Vanity metrics are not a new concept, but they persist because they are psychologically satisfying. Large numbers feel like progress. Follower counts, page views, impressions. These metrics are not worthless, but they are dangerous when they are treated as evidence of business performance rather than evidence of reach or visibility.

I judged the Effie Awards, which evaluate marketing effectiveness, and one of the most consistent patterns in weaker entries was an over-reliance on reach and engagement metrics as proxies for business impact. The strongest entries made a clear, documented connection between marketing activity and commercial outcome. They did not just report that the campaign was seen by a lot of people. They showed what those people did next, and what that meant for the business.

The discipline required to make that connection is not technical. It is intellectual. You have to be willing to ask whether the number you are celebrating actually means anything, and to be honest when the answer is “not really.”

If you want to go deeper on building measurement systems that hold up to that kind of scrutiny, the marketing analytics hub covers attribution, GA4 implementation, and the broader frameworks that make individual KPIs meaningful rather than decorative.

Using GA4 to Track KPIs Properly

GA4 changed the measurement landscape significantly. The shift from session-based to event-based tracking was not just a technical update. It was a different philosophy of measurement. Sessions were a proxy for behaviour. Events are behaviour. That distinction matters when you are trying to build KPIs that reflect what people actually do rather than how long they spent on a page.

The practical implication is that GA4 requires more upfront configuration to generate useful KPI data. You have to define what events matter, set up conversions deliberately, and think about the user experience you want to measure before you start collecting data. Going in without that preparation produces a dataset that is technically comprehensive and practically useless.

Preparing for GA4 properly means mapping your measurement objectives to the event taxonomy before you configure anything. What actions on your site or app constitute meaningful progress toward a business goal? Those are your conversion events. Everything else is context.

One area where GA4 adds genuine value for KPI tracking is A/B testing integration. Being able to connect A/B testing in GA4 to your defined conversion events means you can evaluate whether changes to your site are moving the metrics that actually matter, rather than just the ones that are easy to measure. That is the difference between optimisation that improves KPIs and optimisation that improves business outcomes.

How Many KPIs Is Too Many?

There is no universal answer, but there is a useful heuristic. If your KPI list cannot fit on a single page without requiring footnotes, it is too long. If you cannot recite the three most important ones from memory, there are too many to be actionable.

At a business level, most organisations need no more than five to seven primary KPIs. At a channel level, three to five. Below that, you have supporting metrics that provide diagnostic context but should not carry the same weight as a true KPI.

The discipline of reduction is harder than it sounds. Every stakeholder has a metric they care about. Every channel manager has a number that makes their work look good. Cutting the list requires making explicit choices about what the business is actually trying to achieve, and those choices can be uncomfortable when they imply that some activity is not as important as previously assumed.

When I was managing large ad spend portfolios across multiple clients, the accounts that performed best over time were consistently the ones where the client had the clearest sense of what one or two numbers actually determined success. Not because we ignored everything else, but because having a clear primary KPI forced honest conversations about whether the work was delivering. Accounts without that clarity tended to drift toward optimising whatever was easiest to improve, which is rarely the same as what matters most.

Reviewing and Updating KPIs

KPIs are not permanent fixtures. A metric that was the right indicator for a growth-stage business may be the wrong one for a business focused on retention and profitability. Business objectives change. Market conditions change. The measurement framework needs to keep pace.

A quarterly KPI review is a minimum. The questions to ask are: Is this metric still connected to a live business priority? Has the relationship between this metric and the outcome it predicts held up over time? Are we making decisions based on this number, or just reporting it?

That last question is the most revealing. If a KPI is consistently reported but never acted upon, it is not a KPI. It is a reporting habit. Habits are comfortable. They are also expensive when they consume analytical capacity that could be directed at something that actually drives decisions.

Organic search KPIs are a good example of metrics that need regular recalibration. The relationship between keyword rankings and traffic has shifted as search behaviour has changed. Tracking keywords in Google Analytics alongside traffic and conversion data gives you a more complete picture than rankings alone, but even that picture needs to be interpreted in the context of current search behaviour rather than assumptions formed three years ago.

The Honest Case for Fewer, Better KPIs

If I could retrospectively change one thing about the measurement culture in every agency and client organisation I have worked with, it would be this: the willingness to have fewer KPIs and be genuinely accountable to them. Not to track fewer things, but to be honest about which numbers actually determine success and to stand behind them even when they are uncomfortable.

Most measurement problems are not technical. The tools exist. The data exists. The problem is the human tendency to select metrics that reflect well on the work rather than metrics that reflect honestly on the outcome. Fixing that does not require better software. It requires the commercial courage to ask whether the numbers you are celebrating would survive scrutiny from someone who had no stake in the answer.

That question is worth asking regularly. The businesses and marketing teams that ask it consistently tend to make better decisions, waste less budget, and build a more honest relationship between marketing activity and commercial performance. Which is, in the end, what measurement is supposed to be for.

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