© 2025 Great State

WHAT DO WE MEAN BY BAD DATA?

Let’s clear one thing up straight away. Bad data does not mean incomplete datasets, inconsistent values, or messy, duplicate-filled sources. Data quality still matters.

I’m talking about the metrics that give you bad vibes: the numbers that highlight friction, failure and frustration. These are the KPIs people like to bury in appendix slides: crashes, exits, drop-offs, abandonments, complaints.

They are uncomfortable because they appear to tell you what’s sinking, not what’s winning. But if your goal is effortlessness, this is exactly where the richest clues live. Before we get into which badness to embrace, we need to talk about why your “good” data is not enough on it’s own.

THE LIMITS OF GOOD DATA

Most digital reporting leans heavily on a familiar set of comfort metrics:

• Time on site or in app

• Click-through rates

• Number of interactions

• Page or screen views

These are useful for tracking performance trends. They can tell you whether a campaign drove more traffic, whether a UI change increased interaction, or whether content is “engaging”. They are, to put it bluntly, the fluffy slippers of analytics: well-worn, familiar and safe.

But they are poorly aligned to an era of digital minimalism, which is one of the big themes emerging from Shifting States. The best experiences are no longer the ones that maximise engagement; they are the ones that give people their time back.

If users want to “get in, get what they need, and get out”, then session durations shrinking and interactions declining can actually be a sign of success. If you are only staring at engagement metrics, you will experience this as a plateau - or a problem - rather than as a signal that users are achieving their goals faster.

In short: comfort metrics tell you how much activity happened. They do not tell you how much effort was involved.

THE POWER OF BAD DATA

Bad data, by contrast, is anything that traditionally indicates something is not working for customers. It highlights friction: crashes, exits, abandonments, complaints, dead-end interactions.

These are not fun to report. They can often trigger defensive conversations. But when you reframe them, they become the most valuable inputs into designing effortless journeys.

Ninety per cent of 18–70-year-olds tell us that when price and quality are equal, effortlessness often becomes the deciding factor at the crucial conversion stage.

I look at data every day, but it is tempting to assume that designing effortless experiences is the job of UX and design teams. They are instrumental, of course. But data teams are just as critical. We are constantly asked to surface actionable insights - and when the task is “help us understand what makes this journey effortless”, it can be hard to know where to start.

The answer is counter-intuitive: you get there faster by embracing your “bad” data.

Adrian Bryers

Lead Strategist

So what are we saying. In short, if good data is your comfortable pair of slippers, your bad data is that fancier, more demanding pair of shoes. You will not wear them every day - but when you do, they help you step forward with purpose.

CONTEXT IS EVERYTHING

Of course, no metric good or bad means much without context. Any decent data practitioner already knows this, but it is particularly important when working with friction metrics.

Take a classic bad metric: basket or cart abandonment. On the surface, a rising abandonment rate looks like failure. But what if your digital experience is actually nudging people into a more complex omnichannel journey?

Perhaps customers start a purchase online, then go in-store to see the product in person, and complete the sale there. If you only look at online abandonment, you will assume your journey is broken. If you combine that with third-party shopper panels or in-store data, you may discover that your digital experience is a powerful research tool driving high-value offline sales.

When you report bad data, you must therefore ask:

• What might this behaviour lead to next?

• Where else might the value be realised (another device, another channel, offline)?

• What other datasets - first-party or third-party - could complete the story?

• Bad data without context is misleading. Bad data with context is a blueprint.

HOW TO WORK WITH BAD DATA

Bad data, by contrast, is anything that traditionally indicates something is not working for customers. It highlights friction: crashes, exits, abandonments, complaints, dead-end interactions.

These are not fun to report. They can often trigger defensive conversations. But when you reframe them, they become the most valuable inputs into designing effortless journeys.

Prioritise the metrics that expose friction
Actively seek out the numbers that show pain – crashes, complaints, exits, abandonment, dead ends. That is where effort lives.

Bad data doesn’t mean dirty data
Cleansing, deduplication and quality checks still matter. You cannot design effortless experiences on unreliable inputs.

Strip out ego stats
Vanity metrics cloud judgement. They often reward noise, not value. Remove them from the headline view when the question on the table is effortlessness.

Always report friction contextually

Pair bad metrics with:

•The key action the user was trying to take

•The next step in their journey (if known)

•The customer segment or lifetime stage (new vs seasoned users)

This turns raw negativity into nuanced insight.

HOLD THE LINE ON FRICTION

Effortless digital experiences are not built on comfort metrics. They are built on the insights that reveal effort: where people stall, where they back out, where they get confused, where they simply give up. Stakeholders naturally gravitate towards big green numbers. Part of our job is to show why addressing the skeletons in the closet will have a disproportionate impact on loyalty and conversion.

If effortlessness is becoming the deciding factor in brand choice, then bad data is no longer something to hide in the appendix. It is the most important data you have.