Before you act on your data, consider the context
A number can tell you what happened. It takes context to understand why it happened and that difference matters more than most of us realise.
Think about the kind of data most workplaces rely on every day: staff turnover, engagement levels, productivity, customer satisfaction, complaints, sales figures and progress tracked over time. None of these numbers exist in isolation. Every single one of them is shaped by people and people don't behave the same way in every situation. They respond to the conditions around them: how much pressure they're under, what's changed recently, what they've been told and what they've experienced. So when we look at a metric moving up or down, we're really looking at a reflection of how people reacted to their circumstances at that particular moment.
When the numbers drop, what's the real story?
Picture this: your latest employee engagement survey shows a drop compared to last time. The easy conclusion is that morale is slipping or that leadership needs to do a better job communicating.
But before jumping there, it's worth asking a few more questions. Was there a restructure recently? Has the workload gone up? Did a flexible working policy get scrapped? Was the survey sent out right after a tough announcement?
The results themselves might be completely accurate, a genuine snapshot of how people were feeling at that moment. But without knowing what else was happening, we can easily land on the wrong explanation.
Good news can be misleading too
It's not just bad results that get misread, good ones can too.
A jump in sales might look like proof that a new strategy worked, when really a competitor just left the market. Higher satisfaction scores might get credited to a new wellbeing programme when the bigger driver was a shift in local transport or employment conditions. Even a drop in customer complaints, which sounds like a win, could simply mean people found it harder to actually lodge one.
Data only ever captures a slice of what's going on. It doesn't capture the full context the story played out in.
Why "before and after" comparisons aren't enough
This is exactly why comparing results over time (this quarter versus last quarter, this survey versus the previous one) isn't enough on its own. It's tempting to assume that whatever change we made right before the numbers moved is the reason they moved. But that's correlation wearing the costume of causation.
In my experience, most organisations are far better at recording what changed in their results than they are at recording what else was happening at the time. Dashboards and reports are great at showing movement but they rarely show the full backdrop. They’re designed as if everything outside the metric stayed perfectly still. It never does.
Every team, every workplace, sits inside its own set of shifting conditions: economic pressures, policy changes, staff coming and going, new technology, changing expectations and even local events disrupting the normal rhythm of things. None of this makes the data wrong. It just makes interpreting it a lot more complicated than a single chart can show.
The questions worth asking before you act on data
That complexity isn't something to avoid. It's something to sit with and work through. A useful habit is to pause before reacting to any result and ask:
What was happening around the time this data was collected?
What changed during the period being examined?
Who was affected, and might different groups have experienced it differently?
What outside factors could have played a role?
What's a plausible explanation, even if it's not a comfortable one?
What extra information would help test these possibilities?
This is where qualitative input earns its keep. Conversations, observations, and open-ended survey feedback often show you what the numbers alone can’t — not just what people did but what they were reacting to and why.
Don't let context become an excuse
There's a real risk here worth naming: context can be misused. It's easy to turn it into a get-out-of-jail-free card for an uncomfortable result. "The timing was off." "That group was an outlier." "There was just too much going on." Sound familiar?
Looking for context isn't about explaining away a bad result. It's about resisting the pull towards the first explanation that happens to be convenient. The goal isn't to soften accountability. It's to sharpen the quality of the conclusions we draw so the decisions that follow are actually the right ones.
Bringing the numbers and the story together
When we chase the numbers alone and strip human behaviour away from the conditions shaping it, we risk fixing the wrong problem entirely. We end up blaming one thing when something else was actually driving the outcome.
Data visualisation makes reality more visible. It puts patterns and trends in front of us in a way plain numbers can't. But context is what helps us understand what actually made those patterns possible and what's genuinely needed next. Neither one gets you the full picture alone. The real value comes from putting them together - reading the data and asking what surrounds it.
FAQ
Why is context important when interpreting data?
Context is what turns a number into a story. It explains what was actually happening at the time, who it affected and what else might have played a role. Without it, you’ll easily mistake correlation for causation, and that’s when you end up fixing the wrong problem.
What is the difference between data visualisation and data interpretation?
A good chart makes patterns visible. But interpretation is asking why those patterns are there in the first place. A dashboard can show you the numbers moved, but it takes context to understand what actually made them move.
Why do employee engagement scores drop even when nothing seems wrong?
Engagement scores are sensitive to all sorts of things beyond day-to-day morale: a recent restructure, a policy change, an increased workload and even the timing of the survey itself. That’s why it’s worth asking what else was happening at the time. It helps you separate what’s actually going on from what you assume is going on.
Can data be accurate but still be misleading?
Absolutely, and I see it happen all the time. A rise in sales, a drop in complaints or an improved satisfaction score. They're all accurate, but they can all be caused by things that you have nothing to do with. Without context, you’ll credit yourself for changes that would have happened anyway.
How do you avoid jumping to the wrong conclusion from data?
Pause before you act. Ask what changed during that period, who was affected and what other explanations are plausible – even the uncomfortable ones. Then test your first instinct against what people actually tell you. That’s when you’ll know if your first explanation actually holds up.
Is using context to explain data the same as making excuses?
No — there's an important difference. Making excuses means dismissing a result you don't like. Using context means testing whether your explanation is actually correct before acting on it. The goal is better decisions, not lower accountability.
What questions should you ask before acting on a data result?
Start with these. What was happening when this data was collected? What changed during that period? Who was actually affected? What else might have played a role? And what would help you confirm whether your first explanation is the right one?
From Numbers to Insight: Work With Selena
Understanding data means understanding context. When your team is ready to move beyond the numbers to the story behind them, I run keynotes and workshops for organisations ready to shift how they read evidence. You can find my current lineup of keynotes and workshop topics at selenafisk.com/speaking.
