We've all been there. You're staring at a spreadsheet full of numbers, trying to turn it into something clear, compelling and easy to understand. But choosing the right chart, spotting the important patterns and making everything look presentable can be surprisingly time-consuming. Luckily, AI can help.

In Apolitical's recent event AI Skills Lab | AI for Data Visualisation, data and behavioural scientist Ali Seyhun Saral demonstrated how tools like Microsoft Copilot and Excel can help public servants clean, analyse and visualise data, all while saving time and making insights easier to understand.

Here are six practical takeaways to help you get started.

1. Clean your data before making it pretty

Even the cleverest AI tool struggles with a chaotic spreadsheet.

Fortunately, AI can help with the preparation of data. Before creating your charts, you can use it to summarise your dataset, explain what each column contains, identify inconsistencies and clean up messy information.

A few habits make the process smoother:

  • Keep your original data intact so you can check or reverse changes.
  • Use a clear structure where each row represents one observation.
  • Create new sheets for AI-generated calculations and outputs.
  • Avoid unusual spreadsheet layouts that could confuse the tool.

A little preparation can save a lot of troubleshooting later.

2. Turn messy text into meaningful insights

Got hundreds of survey responses, feedback forms or open-ended comments sitting in a spreadsheet? AI can help you make sense of them.

One of AI's most useful capabilities is turning unstructured text into data you can analyse. It can identify recurring themes, categorise responses and assess sentiment, transforming large volumes of written feedback into patterns you can measure and visualise.

Information that once required hours of manual sorting can now be analysed and turned into useful charts much faster. This makes it easier to spot trends, understand what people are saying and uncover insights that might otherwise go unnoticed. Just remember to check AI's interpretations for accuracy.

3. Let AI help you choose the right chart

Sometimes a simple bar chart tells the story better than something with a dozen colours and arrows.

AI can help you identify the most effective way to present your data, suggesting different chart types based on what you're trying to communicate. It can even help you create more advanced visualisations, like Sankey diagrams and heat maps, without needing to write code yourself.

The key is knowing your audience. A senior leader might need a quick overview, while an analyst might want the finer details. Offer this information to your AI tool and ask it which format will work best.

4. Start broad, then get specific

Don't expect the perfect chart on your first attempt.

AI works best when you treat it as a collaborative tool rather than expecting a finished product from a single prompt. Start with a general request, then refine your instructions as you go. Ask AI to visualise your data, then request changes to the layout, labels, colours or how the information is organised.

For better results, be specific about which sheets, columns and data ranges you want AI to use.

And if your first chart looks a little questionable, that's fine. It can just be a first draft rather than a finished product.

5. Test your charts on an AI audience

A chart might make perfect sense to you, but will it make sense to everyone else?

One useful way to use AI is to ask it to act as your intended audience. Whether you're presenting findings to senior leaders, colleagues or the public, AI can help you identify confusing labels, unclear messages or information that needs more explanation.

Try asking AI to review your chart from the perspective of someone unfamiliar with the data. What stands out? What's difficult to understand? Is the main takeaway obvious?

This gives you a practice audience and helps you refine your visualisation before sharing it.

6. Remember to use AI responsibly

While AI makes data analysis and visualisation faster, it also changes how we need to check our work. When a tool handles calculations and creates charts behind the scenes, it's important to verify that the results are accurate.

Use checklists, inspect random samples and cross-check totals against your original data. You can even start a fresh AI conversation to independently review the results, rather than relying on the same chat that produced them.

And before uploading any data, check your organisation's AI policies and privacy settings. Never use confidential or highly confidential information unless you have explicit permission.

Want to see these techniques in action? Watch the highlights video below, or visit the event page (opens in a new tab)for the full recording, additional resources and full speaker bios.

This article was written with the help of AI using an anonymised transcript and edited by a human before publishing.

Make sure to share your own thoughts with the author by leaving a comment below