Beyond the Spreadsheet: Alternative Ways to Analyse Data

Ishe Chinyoka

- 5 minutes read

Table of Contents

One of the questions I often receive whenever I talk about plain text is this:

“If you don’t use Excel or LibreOffice Calc, how do you analyse your data?”

It is a reasonable question. For decades, spreadsheets have become synonymous with data analysis. Many people cannot imagine working with numbers without seeing rows and columns on a colourful grid.

My answer usually surprises them.

I do not reject spreadsheets. I simply question whether the spreadsheet is the destination or merely one possible interface.

This distinction changes everything.

Is the Spreadsheet the Goal?

Programs such as Microsoft Excel, LibreOffice Calc, and Google Sheets are remarkable pieces of software.

They provide an intuitive visual environment where you can:

  • enter numbers,
  • apply formulas,
  • create charts,
  • sort information,
  • build reports.

Their greatest strength is that they let you see your data as you work. The familiar grid of rows and columns makes information approachable, especially for beginners.

But ask yourself this question:

What are you actually trying to accomplish?

If your objective is to calculate averages, find trends, clean messy data, merge several datasets, build statistical models, or create repeatable reports, then the spreadsheet itself is only one possible tool.

The real goal is data analysis.

Once you realise this, the world becomes much larger.

Plain Text Starts Earlier

As a textsmith, I rarely begin with a spreadsheet.

Most information arrives as text anyway.

CSV files.

TSV files.

Log files.

JSON.

XML.

SQL exports.

Even spreadsheets can be exported into CSV with a single click.

These formats are portable, version-controllable, and readable without proprietary software. They also fit naturally into the Unix philosophy: treat data as text and build small tools that work together.

Instead of opening a graphical spreadsheet, I often begin with commands like:

cut
sort
uniq
grep
awk
sed
join
paste

These tools may appear simple individually, but together they can clean, transform, filter, summarise, and reshape enormous datasets surprisingly quickly.

When the Data Gets Larger

The moment datasets become moderately large, the spreadsheet begins to show its limitations.

Scrolling becomes slower.

Copy-and-paste becomes error-prone.

Hidden formulas become difficult to audit.

Manual operations become difficult to reproduce.

This is where programming languages shine.

Python: The Modern Data Workhorse

Python has become one of the dominant languages for data analysis, and for good reason.

Libraries such as:

  • NumPy
  • pandas
  • Polars
  • Matplotlib
  • SciPy

allow you to manipulate millions of records using concise, readable code.

Instead of manually repeating operations, you write them once.

Instead of wondering how you obtained a result six months later, your script documents every step.

Instead of clicking menus, you describe your analysis.

This makes the work reproducible.

Perhaps even more importantly, Python can import data from Excel workbooks just as easily as it can read CSV files.

The spreadsheet becomes an input source rather than the working environment.

R: Statistics First

Although Python receives much of today’s attention, I continue to appreciate R.

Unlike Python, which is a general-purpose programming language, R was designed from the beginning for statistics.

Its libraries provide an astonishing range of statistical techniques, visualisations, and modelling tools.

For researchers, scientists, economists, and analysts, R often feels like the natural language of data.

Packages from the tidyverse ecosystem make importing, transforming, analysing, and visualising datasets remarkably expressive.

Like Python, R reads Excel files without difficulty.

Again, the spreadsheet becomes simply one source of data among many.

Reproducibility Matters

One of the greatest advantages of text-based workflows is reproducibility.

Imagine cleaning a spreadsheet manually for three hours.

Next month another file arrives.

Do you repeat every mouse click?

Or do you simply run your script again?

A shell script.

A Python notebook.

An R script.

A Makefile.

These become living documentation of your entire analysis.

Anyone—including your future self—can reproduce exactly the same results.

That is much harder to achieve when most of the work happened through menu clicks.

Version Control Loves Plain Text

There is another advantage that textsmiths quickly appreciate.

Git understands text.

It happily tracks changes to CSV files.

It shows exactly which lines changed.

It allows meaningful comparisons between versions.

Binary spreadsheet formats are much harder to inspect. Git can tell you that the file changed, but not necessarily how it changed.

When your workflow is text-based, version control becomes a natural companion.

Are Spreadsheets Obsolete?

Not at all.

I still use spreadsheets when they are the right tool.

They are excellent for:

  • quick exploration,
  • simple budgeting,
  • presenting tables,
  • collaborating with colleagues,
  • sharing information with less technical users.

In many organisations they remain the common language of business.

The mistake is not using spreadsheets.

The mistake is believing they are the only way to work with data.

Thinking Like a Textsmith

A textsmith approaches data much like any other form of text.

Data has structure.

Structure can be described.

Descriptions can be processed.

Processing can be automated.

Automation can be repeated.

Whether the input is a server log, a CSV export, an Excel workbook, or a database dump, the principles remain remarkably similar.

The spreadsheet is simply one representation of structured information.

It is not the information itself.

Final Thoughts

The question, then, is not:

“How can you survive without Excel?”

The better question is:

“What am I trying to achieve?”

If your destination is meaningful analysis, reliable automation, reproducible research, and scalable workflows, then there is abundant life beyond the spreadsheet.

Spreadsheets remain valuable tools.

But they are one stop along the journey—not the destination itself.

As textsmiths, we learn to look beyond the grid.

We focus instead on the data, the transformations we wish to perform, and the story those numbers are waiting to tell.