Why we believe in Small Data

tl;dr: The power of your laptop is way better than even the most powerful data centres of 10-15 years ago. We've also gotten way better at making efficient, clean software. We're committed to delivering these wins to everyone, making everyday data analysis a modern process.

Every day at 2 AM, somewhere in the world, a brilliant professional is staring at their screen, struggling with Excel. Not because they lack expertise in their field, but because they're trying to wrangle together some messy datasets to prepare for a report, and they can't quite get the right set of formulas together - no matter how relentlessly they prompt ChatGPT!

Sound familiar?

There's a quiet revolution happening in the world of data. While tech headlines scream about artificial intelligence, machine learning, and massive data centres, we at Columns are asking a simpler question:

What if we've overcomplicated everything?

The issue with Big Data

For years, Big Data evangelists have spread the word that "more data = better results". That if you just collect enough data, there will be limitless ROI from the insights.

This leads to an explosion of expensive and complex solutions:

• The hiring of a glut of "data specialists" (engineers, scientists, etc.) just to create pipelines, clean data, and manage infrastructure, which takes resources away from actually solving problems
• Simple analyses made difficult through red tape and data engineering frictions
• Valuable insights getting lost in oceans of irrelevant data
• Teams burning compute costs on data they rarely use

The reality is that the thing preventing organisations from being really "data-driven" usually isn't some red-hot prediction model, machine learning pipeline, or AI implementation.

From my experience, it's usually inefficient processes and poor ways of working with data on the ground floor.

We want to change this.

It's been estimated that spreadsheet users waste at least 20% of their time on repetitive work. And as the sheer size of data that individuals have to work with on a daily basis increases, we wouldn't be surprised if the time wasted increases too.

So how are people really using data?

If you're a domain expert who works regularly with data (that is, not a data expert), you're probably familiar with these scenarios:

• Your data keeps changing (there's some new information that needs inclusion!)
• Nothing is ever simple (every report or cleaning exercise requires reconciling multiple sources)
• The data is messy (think handwritten records, badly made Excel spreadsheets with unclear formulas that need to be merged somehow)

In doing many of these tasks, one is probably finding themselves wrestling with the tool they have access to itself, rather than focusing on their actual expertise.

These aren’t isolated cases or just ‘part of the job’—they’re barriers holding people back from actually solving problems.

The way forward: thinking smaller

One of the key motivations behind Columns is in this simple philosophy: your laptop today has more computing power than entire data centers from a decade ago.

Modern hardware is incredible, yet we've somehow convinced ourselves that every data problem needs cloud infrastructure. Humans are also way more agile and mobile than ever before, and we've all found ourselves somewhere without internet,suddently completely blocked from progress as all your tools moved to the cloud!

Local-first means more than just efficiency — it’s about regaining control over our tools and data. With the right tools, processing data right on your device preserves privacy, enhances security, and reduces the reliance on costly cloud solutions.

Excel and its ilk was invented over 40 years ago to solve the problems of that time, with the tech available to them at that time! We deal with so much more complexity and scale nowadays, that requires real solutions, not dragging down a formula cell.

However, we don't see software addressing the problems of today that also feels like it was built with humans in mind. We believe there's a better way. Observing the Small Data Manifesto, we're committed to building software that lets you:

• focus on the data that actually matters to you
• process it right on your machine, with everything just working out-of-the-box
• view, clean, and analyse data in seconds without coding

We see a future where individuals and teams are empowered to work directly with the data that matters to them.

Instead of needing to defer to code (or worse, needing someone else to code for you), we're building something that lets anyone do data simply.

The Small Data movement is just beginning, but it's gathering momentum. If you're tired of overcomplicated tools, frustrated with cloud dependencies, or just believe there must be a better way to work with data, you're already part of it.

We're building tools for this future - a future where working with data is fast, joyful, and accessible to everyone. We'd love for you to join us on this journey.

---

Keen for a demo or to chat about the future of the Small Data movement? Get in touch - we'd love to chat.