I’ll research current data on analytics usage, dashboard adoption, and data-driven decision-making before writing.I have strong data. Let me get one more angle on dashboard usage and the cost of vanity metrics.I have enough solid, attributable data to write the article.
The dashboard nobody opens
Walk into most mid-market marketing teams and you’ll find a beautifully built GA4 dashboard that gets opened once a month, glanced at for the monthly report, and then closed. Sessions are up. Bounce rate is down. Average engagement time ticked along nicely. Everyone nods. Nobody does anything differently on Monday.
That dashboard is not measurement. It is decoration.
The hard part of analytics was never collecting numbers. GA4 will happily collect millions of data points whether or not a single one of them changes what your business does. The hard part is building a measurement system where each number forces a decision. If a metric can double or halve and your team would do exactly the same thing tomorrow, that metric is taking up space and earning nothing.
This is the problem we see most often on mid-market sites, and it has very little to do with GA4 being difficult. It has everything to do with the order in which decisions got made.
Why your dashboard fills up with noise
The default GA4 reports are built for everyone, which means they are built for no one in particular. Out of the box, you get sessions, users, views, engagement rate, and a tidy set of charts that describe traffic without ever connecting it to whether your business is winning. Those are the numbers that look impressive and feel productive to report. They are also, in the language of vanity metrics, exactly the trap.
The framing comes from Eric Ries in The Lean Startup. A vanity metric is a data point that is easily manipulated and appears impressive on a superficial level, but fails to provide meaningful insight into business performance or inform future strategy, highlighting metrics that are more about stroking egos than driving intelligent action. Page views are the classic example. Imagine the metric suddenly doubles or halves. If your number of Twitter followers doubles overnight, you probably do nothing differently. But if your cost per qualified lead doubles, you have an immediate, specific action: investigate your ad campaigns, landing pages, and targeting.
That is the test. Most GA4 dashboards fail it because they were assembled from whatever the tool surfaced by default, not from the questions the business actually needed answered.
And this is not a fringe complaint from a few purists. According to a study by Viant, 36% of CFOs cite the use of vanity metrics by CMOs as a top concern, reinforcing the perception of marketing as a cost center rather than a growth engine. When marketing can’t prove its value in dollars and cents, budgets get cut and credibility erodes. Your dashboard full of sessions and pageviews is not just useless internally. It is actively undermining how the rest of the business sees marketing.
The real culprit: tracking added after the fact
Here is the uncomfortable pattern. A company launches or relaunches a website. The build is done, the site is live, everyone is relieved. Then someone says “we should set up analytics,” and a tag goes on, and GA4 starts collecting whatever it collects by default. Nobody defined what success looked like before the site shipped. So the dashboard reflects what the tool measures automatically, not what the business cares about.
That sequencing is the original sin. When measurement is bolted on at the end, you get data that describes activity but never answers a question. The form on the contact page fires a generic page view but not a structured event you can tie to pipeline. The demo request and the newsletter signup both count as “conversions” with no way to tell the high-intent action from the low-intent one. Three months later someone asks “which channel actually drives demos?” and the honest answer is that the data was never set up to know.
In our projects, we define measurement requirements during prototyping, before a line of production code is written. What counts as a meaningful action on this page? How will we know if this page is doing its job? Which events need to fire, with what parameters, so that six months from now a marketer can answer a real question without rebuilding the whole setup. That is the difference between launching with visibility and launching with a backlog of “we should be tracking that.”
GA4 didn’t help, but it isn’t the villain
The platform migration genuinely shook teams. The fundamental shift from Universal Analytics’s session-based model to GA4’s event-based architecture represented more than a technical upgrade. It was a complete reimagining of how to measure and understand user behaviour. A lot of teams rebuilt their tracking in a hurry, copied old habits where they could, and never stepped back to ask what they actually wanted to measure.
Adoption itself is no longer the issue. GA4 is now the default. Current data shows that over 15 million websites use GA4, making it the de facto standard for web analytics regardless of individual opinions about the platform. The Influencer Marketing Hub’s Digital Marketing Benchmark Report 2024 found that among the companies that have transitioned to GA4, 40.9% reported a complete understanding and feel comfortable using GA4, while 39.5% are still learning the platform’s features. So roughly two in five teams using the tool are still not confident in it. That confidence gap is exactly where unused dashboards breed.
What a useful number looks like
The opposite of a vanity metric is not a complicated metric. It is one with a direct path to a decision. If you have to make several logical leaps to form a connection, more likes might lead to more brand awareness, which could eventually lead to more sales, it’s likely a vanity metric. An actionable metric has a direct path. For example, demo request form submissions directly impacts the sales pipeline.
Context matters here, and this is where teams overcorrect. A metric is not vanity or actionable in the abstract. It depends on who is using it and for what. When it comes to KPIs, context is everything. The metrics you use depend on at what level and for what purpose you are using them. Social media engagement rate is usually considered a vanity metric because it is not connected directly to sales, but for a social media manager, it can be an actionable metric.
So the question is never “is this a good metric.” The question is “whose decision does this number change, and what would they do if it moved.” If you cannot name the person and the action, cut it from the dashboard.
For most B2B mid-market sites, the numbers that pass that test cluster around a handful of things:
- Qualified lead actions, separated by intent. A demo request and a gated PDF download are not the same event and should never share a conversion count.
- Conversion rate by source, so you can see which channels send traffic that does something rather than traffic that simply arrives.
- Cost per qualified lead, which gives finance a number in the language they respect.
- Drop-off points in a defined funnel, so when conversions fall you know where, not just that.
Notice what is missing. Total sessions. Total users. Pageviews. Those belong in the engine room, useful for diagnosis, dangerous on the headline dashboard because they invite the team to feel good without learning anything.
The cost of a dashboard nobody trusts
An unused dashboard is annoying. A distrusted one is expensive. When two people pull the same report and get different numbers, the whole exercise collapses. When leaders are able to pull different reports based on different numbers, confidence evaporates. Major business decisions get derailed because teams can’t agree on basic metrics. Without consistent, and accessible, information, even the most sophisticated analytics become useless.
This is not a rare edge case. BARC’s research on decision-making found that the quality of data, cited by 40% of respondents, is the second most common barrier to data-based decision-making, reflecting a need for more attention to be paid to data quality and more efficient data governance. And when the data isn’t trusted, people quietly fall back on instinct. The same research found that 39% of companies state that gut feel and experience is good enough to make decisions with no relevant information.
Think about what that means in practice. You paid to build a measurement system. Nearly four in ten teams are making calls on gut feel anyway, often because the system gives them numbers they don’t believe. The dashboard isn’t just neutral clutter at that point. It is a sunk cost that has trained your team to distrust data altogether.
The opposite failure mode is just as real. Some teams drown. Companies can get stuck analysing endless information. If every dashboard drives more questions than answers, that’s a losing battle. Businesses chasing endless data or sifting through a deluge risk delaying critical decisions while their competitors move ahead. Analysis paralysis is real and costly. A dashboard with sixty widgets is not more rigorous than one with six. It is usually a sign that nobody was willing to decide what mattered.
Start from the decision, not the data
The fix is a reversal. Most teams start with the data they happen to have and try to wring insight out of it. That is backwards. MIT Sloan’s research on analytics programmes is blunt about this. Don’t aimlessly follow data. Anchoring initiatives on available data often leads to a focus on the wrong questions and can reinforce preexisting beliefs. The researchers argue that staying focused on available, mostly historical data often leads decision-makers down the wrong path, gaining insights to questions that don’t address the fundamental business problem.
The payoff for getting this effort to actually connect to outcomes is not guaranteed. Accenture’s survey, cited by MIT Sloan, found that only 32% of companies reported realising tangible and measurable value from data. Two-thirds are spending money on analytics and getting little back. The common thread among the third that succeed is that they started with the decision and worked backwards to the data, rather than the reverse.
What this looks like for a website project: before you choose a single metric, write down the three or four decisions your team will actually make in the next year. Where to spend the next marketing pound. Which page to rebuild first. Whether the new pricing page is pulling its weight. Then ask what evidence would change each of those decisions. That list becomes your dashboard. Everything else is reference material at best.
One more trap: numbers without the why
Even a clean, actionable dashboard has a limit. It tells you what happened, rarely why. Companies can’t over-rely on analytics without human context. Data can reflect what happened and may predict what happens next, but it frequently misses the why. A funnel report might show a sharp drop-off at the pricing step. It will not tell you whether the prices scared people off, the page loaded slowly, or the call-to-action was confusing. That is where qualitative work, session recordings, user testing, and actually talking to customers, has to sit alongside the numbers.
The strongest teams pair the two without letting either dominate. They bring the dashboard into the room where decisions get made, then they argue about what it means with people who understand the business, not just the data. A number that prompts a real discussion is worth ten that prompt a polite nod.
What to do this week
You don’t need a replatform to fix this. You need an afternoon and some honesty. Here is where to start.
- Audit your current dashboard against the doubling test. Go widget by widget. For each one, ask: if this number doubled or halved tomorrow, who would do what? Delete every widget where the honest answer is “nothing.”
- List your real decisions. Write down the three or four choices your team will genuinely make this year about the website and marketing spend. These define what your dashboard should contain.
- Check that your conversions are separated by intent. If a high-intent action like a demo request shares a conversion count with a newsletter signup, fix that first. It is the most common and most damaging setup error we find.
- Verify two people get the same number. Have two colleagues pull your headline metric independently. If they disagree, you have a definition problem, and no amount of dashboard polish will fix the trust gap underneath it.
If you’re planning a website project or staring at a dashboard you’ve stopped trusting, the cleanest moment to fix measurement is before the next build, not after. That is exactly the work we structure into our prototype sprints, and you can book your free discovery call to talk through your own situation.
The goal is not a prettier dashboard. It is a shorter one, where every number earns its place by changing what you do. Strip out the noise, and what’s left will finally get used.
Ready to get started?
