What makes a dataset worth telling a story about?
“How do I know that my data insight is interesting?”
“How do you know where to look for cool data? Global or economic data can be so boring.”
These are just some questions I’ve received in my career as a data storytelling professional. This shows that one of the first hurdles most people face in data storytelling is picking the right data.
They stem from a belief that there are some datasets that are worthy, while others less so based on their topic coverage. Data concerning subjects like GDP or Purchasing Power Parity, for example, may be seen as less exciting and less worthy of a full-length data story.
As a result, many of us looking to start our data storytelling journeys spend hours searching for an interesting dataset that we believe will inspire us to tell great stories. This sentiment suggests that datasets are scattered across the world wide web, and only a few gems are waiting to be discovered. Without the right dataset to inspire us, we find it tough to get started.
I believe this to be a misconception. What makes a dataset worth telling a story about is not determined by the nature of the data, but from the questions that we ask of it. Depending on the query that we have, even the most mundane of datasets could be worth looking into.
Exactly what determines your dataset’s storytelling potential? We pick out a few key framings to consider in your query.
Some of the best stories are hiding in plain sight. One of our favourite stories at Kontinentalist is about instant noodles — arguably, a delicacy well-loved and frequently consumed by people all across Asia. Our query started with a simple question: why are some instant noodle packets larger in size than others? What goes into the decision-making process of manufacturing these noodles? The result of that was a fun investigation into the world of noodle types, flavours, and variety that later on inspired other similar works.
A good data storyteller is able to look at what appears boring, and recognise there is something to be examined. This could be for something as simple as the location of lampposts in Singapore, and a query about which region has the most or the least.
Everyone loves a piece that challenges conventional wisdom or beliefs. Some of the best data stories available investigate a belief that people hold and ask, “is that actually true?” Mythbusting makes a story memorable and controversial, because it forces your readers to rethink what they know, and discuss their assumptions.
For example, for a long time, Chinese cuisine was held up as the prime example of unethical shark consumption, specifically shark fin. This issue was famously covered in a documentary by Gordon Ramsay.
However, a close look at the data tells a different story. In a 2019 report by TRAFFIC, Brazil came up as the top importer of shark meat between 2008 to 2017 with a wide lead, followed by Spain. In 2021, WWF reported that the EU accounts for 22 percent of global trade in shark meat, with Spain being the top exporter and Italy the top importer. Kontinentalist also briefly reported on this back in 2019.
Fishery data may not appear the most exciting, but the numbers reveal a surprise finding that topples conventional narratives.
Data is often more than just spreadsheets. While numbers can seem abstract, they ultimately describe people, the world, and everyday life. Something I’ve noticed recently is people’s strong reaction and rejection to the latest reports of household median incomes in Singapore. It illustrates that numbers are never neutral, and they come with preconceived notions of access, fairness, and equality that makes many upset.
At the start of 2026, the Department of Statistics in Singapore released their latest numbers for Singapore’s median household income, which crossed $12,000 for the first time. This was shared alongside other metrics, such as the lowest inequality since 2015. The Straits Times shared this news in a graphic, but the Instagram post is full of comments that are doubtful, distrustful, and in disbelief of the numbers. This goes to show that numbers are not neutral, and can trigger a wide range of emotions in what they represent.
The best stories make numbers relatable and concrete. They do not simply state that “incomes are rising”, but attempt to break it down that makes sense for the everyday person, such as what affordability may mean to a single mother with two children while working full-time. In a story I wrote about housing in Singapore back in 2022, we explored what housing affordability means to unconventional families.
As data professionals, we spend a lot of our time exploring the spread of a dataset, namely in the minimum, maximum, or middle parts, and what those represent. We rarely talk about where it might be erroneous, broken, or where everyday life defies expectations – the outlier. The straggler, random, seemingly erroneous data point that stands out.
Some of my favourite pieces in data journalism have been about outliers. For example, in this story about India’s polling stations by Reuters, they explore the lengths in which India goes to ensure that their citizens not only have the right to vote, but to access it as well. In a key graphic, it highlights exceptional locations, such as a temporary booth set up for one voter in remote mountains, or the highest polling booth. Likewise, one of my favourite artists, Mona Chalabi, won her Pulitzer for a story imagining Jeff Bezos’ insane wealth with illustrations.
If you’re stuck with a dataset that you’re not keen on, and all other considerations fail, there’s always a story to be told about what’s changed. Not every data story has to be focused on a key takeaway or insight, and many are simply a journey through time.
There are a ton of examples out there that do this well, but for illustrative purposes, we look at an example we did at Kontinentalist on queuing in Singapore. Despite being an activity we love to hate, Singaporeans are culturally known for our avid queuing, from iPhones, to National Day Parade Tickets, to Hello Kitty toys. Instead of taking these as one-off episodes, we looked at the country’s history of queuing and tracked its highs and lows, looking for patterns that tell us more about ourselves.
While data can appear authoritative, accurate and correct, it is in fact often riddled with human errors. Data can have missing rows, incorrectly entered information, or signs of an intentional cover-up. Looking deeply into what’s unavailable in the dataset can reveal interesting questions about what’s included or not.
At Kontinentalist, we applied this specific framing in a dataset about trees in Singapore for an artwork commissioned by the Singapore Art Museum, titled the Hidden Green. We looked beyond the dataset to understand what greenery may not be captured in a spreadsheet, but shows up in other ways, be it satellite imagery or through the cracks of concrete sidewalks. This revealed to us how nature is administered, manicured, and carefully curated for various urban needs.
These pointers tell us one thing: the dataset itself is rarely ever the story, but it is in the questions you ask of it. From boring polling stations, to lampposts, each dataset can reveal multiple narratives depending on our framing. Good data storytellers don’t hoard data, they ask the right questions, an ability we all have. These six suggestions are meant to provide you with a good starting point, but the rest comes down to patience and persistence. So the next time you chance upon a dataset that may appear simple – do not give up, and keep digging!
Header photo by Maxim Berg on Unsplash
Lapis is created by Kontinentalist, an award-winning data storytelling studio based in Singapore.