A business owner who looks at a GA4 dashboard showing "Direct traffic: 60% of conversions" without realising this category is inflated by missing UTM tags misattributing actual campaign traffic believes they have hard data proving something that the data does not actually show. This false certainty is more dangerous than acknowledged uncertainty, because it closes off the curiosity that would otherwise prompt further investigation.

Bad data does not announce itself as bad. A dashboard showing numbers looks identical whether those numbers are accurate or distorted by tracking gaps, misconfiguration, or attribution errors covered elsewhere in this pillar (see analytics mistakes that mislead your decisions and where data gets lost in your tracking setup). The visual authority of a chart or number creates confidence that the underlying methodology may not actually support.

This is not a case for avoiding analytics in favour of pure intuition it is a case for treating analytics with appropriate rigour: verifying it is actually configured correctly before trusting it (see how to set up GA4 the right way), understanding its specific limitations and gaps, and being willing to say "we do not actually know this yet" rather than presenting an unreliable number with false confidence.

When these conditions are met, data consistently outperforms intuition it is not biased by recent memorable events, does not suffer from confirmation bias toward a preferred channel, and can reveal patterns across volumes of activity no individual could mentally track. The point of this piece is not that data is inferior to intuition in principle, but that incomplete or broken data is genuinely worse than honest uncertainty, and getting analytics right is worth the investment specifically because good data is so much more valuable than bad data dressed up to look authoritative.