*From the “Data but Make it Personal” series – Part 2
Why zooming out matters in life and in data
The past five weeks of training have been packed – and I mean packed – with information. We have explored a whole range of data analysis and visualisation tools: building dashboards in Tableau with Level of Detail expressions (and the ever-frustrating containers), cleaning and transforming data in Alteryx, writing DAX measures and building data models in Power BI…the list goes on.
We’ve gained skills at a pace I wouldn’t have thought possible. And the interesting thing is, I’m only able to see that now because I took a moment to zoom out – probably the four days of public holidays helped a little too.
In the day-to-day, it hasn’t felt like progress. There have been days when I’ve left work carrying not only my computer, but also a very heavy imposter syndrome in my bag. Most of the time all the information just felt like noise – like living at the row level of a dataset that keeps growing faster than I can keep up.
From Chaos to Meaning
When working with data, it is important to start at the row level. You need to identify the granularity of the dataset and actually see what you’re dealing with. Row-level views also help when performing cleaning tasks, and they’re a good way to confirm that the calculations you’re making are returning the correct values.
However, if you want to extract something truly meaningful from all the data, you have to aggregate.
Aggregation is about stepping back. It’s about realising that one frustrating moment (or one row with nulls) doesn’t define the whole dataset, or the whole experience.
In the thick of training, every lesson, every error message felt like a separate, disconnected event. But zooming out, I can see the pattern: growth. Learning. New skills stacking on top of each other, even if it didn’t feel like it at the time.
Maybe, what aggregation does for data, it also does for life.
So, let’s talk about aggregation and why it can be the key to making sense of both datasets and busy weeks.
What is Aggregation, anyway?
In simple terms, aggregation is the process of summarising detailed data to make it more understandable – and more useful.
When we aggregate, we move away from the individual row level and start combining data points into bigger, more meaningful numbers.
Instead of obsessing over each transaction, we ask questions like:
- What was the total sales revenue this month?
- What was the average time spent in support calls last quarter?
- How many clients visited each store this year?
Aggregation gives structure to chaos. It’s what turns 100,000 rows of raw data into a few clear KPIs on a dashboard.
Common Types of Aggregations
Here are the most common types you’ll see (and use all the time):
- SUM – Add up all the values in a field. (Total revenue, total hours worked, total steps walked).
- AVG (Average) – Find the mean value across all rows. (Average products per order, average time spent on a website).
- COUNT or COUNT DISTINCT – Count how many rows exist. (Total number of customers, total number of orders).
- MIN and MAX – Find the lowest and highest values. (Maximum date in the dataset, minimum temperature recorded over a certain period).
Each of these functions helps summarize mountains of data into a few actionable insights.
Grouping Before Aggregating: An Important Step
Aggregation becomes even more powerful when it’s combined with grouping. Sometimes, you don’t just want one total, you want to see totals by category. For example:
- Total sales per region
- Average billing per client
- Number of tantrums per bedtime routine (just hypothetically… maybe)
In data tools, when you GROUP BY, you are telling the system: “Hey, before you sum (or count, or average), split the data into groups based on a specific field”.
This helps to take that big number and turn it into insights:
- Which region is driving the most sales?
- Which projects are taking the most time?
- Which days are going suspiciously well?
Grouping before aggregation helps you uncover patterns hidden inside the data – and make smarter, more targeted decisions.
The Limits of Aggregation
Of course, it isn’t always the perfect solution. Like everything in life, it comes with trade-offs.
When you summarise data, you lose detail – and sometimes, those details matter:
- Averages can hide extremes. A few really bad customer reviews can get buried inside an overall 4.2-star rating.
- Adding totals might miss that some transactions are suspiciously high or low.
- Aggregating after-work hours of blogging and dashboarding over a month might sound great… until you realise all of them are crammed into the last few days and you haven’t slept more than four hours a night.
That’s why good analysts don’t just aggregate blindly.
Good analysts zoom in and out – going between the row level and the summary – making sure the big picture actually reflects reality.
Because sometimes the outliers are the story.
Seeing the Big Picture
Learning to aggregate – in data and in life – is about recognising that meaning often lives above the row level.
Not every day will be perfect.
Not every moment will sparkle.
But when you zoom out, when you group and summarise, the bigger picture tells a much richer story.
The aggregated view usually shows you’re doing better than you think.
