How to turn any dataset into a Data Story for journalists

You have a spreadsheet from a press office, a FOIA request or an open-data portal: 3,000 rows, 14 columns and a deadline. Getting a story out of it usually means an afternoon of pivot tables, then a chart tool, then a fight with the CMS.

Datastripes is a browser tool that shortens that afternoon. It reads the spreadsheet and drafts a data story from it, with headline findings, charts and the paragraphs that explain them, and you edit the draft the way an editor would. You still do the reporting. The tool takes the first pass through the data so you have more time for calls, context and copy.

The result

Press enter or click to view image in full size

What you need

• A CSV, Excel (.xlsx) or Google Sheets file. It works best with one row per record, such as one row per country per year, per school or per contract.

• A browser. There’s nothing to install and no formulas to write.

• About ten minutes.

Your file stays on your device while you explore, and nothing goes public until you publish it.

Press enter or click to view image in full size

Step 1: Drop in the file

Go to datastripes.com, create a project and drag your spreadsheet in.

Datastripes reads every row and works out what each column holds: numbers you can measure (spending, cases, votes), categories (country, party, region), dates and years, and IDs to ignore. It recognises country names too, so country comparisons work without extra setup.

You don’t label anything by hand. If it misreads a column, you’ll see it in the draft and can leave that chapter out.

Step 2: Read the first draft

A few seconds later you have a draft. Each chapter answers one question about your data with a chart and a short finding in plain English:

• Who leads and who lags on a measure

• The year a trend changed direction

• How the same entities compare before and after that turning point

• Whether a few entries account for most of the total

• Whether two measures move together, such as spending and outcomes

Read it like a stringer’s first file. Some chapters will be your story and some will be noise.

Step 3: Direct the story

Open “Direct this story”. It has three steps.

First, write the headline and the conclusion you want readers to take away.

Second, edit the chapters. Keep the ones that matter, delete the rest and put them in order. To add a chapter, pick from questions phrased with your own column names, like “Which regions grew the most in renewable share?”, or follow up on an existing one with “those same regions, before vs. after 2015”.

Third, choose a format and your outlet’s colors (both covered below).

It checks your conclusion against the data

Most AI chart tools will illustrate whatever you tell them to, which is a risk in a newsroom. When you write your conclusion in Datastripes, it lists the chapters that support it, and if the data only partly backs the claim, it tells you before you publish.

It also avoids a few common spreadsheet mistakes:

• It doesn’t add up rates or averages into totals.

• It leaves out years with only a handful of records, so they aren’t presented as a trend.

• It doesn’t double-count cumulative columns as if they were yearly figures.

You still need to know where the data came from, what it leaves out and who collected it. The check makes it less likely that you publish a number that falls apart at the first follow-up question.

Step 4: Publish it in the format your readers use

The same story can go out in eight formats:

FormatGood forArticleA classic scrolling explainerSlidesBriefings, talks and newsroom meetingsChatA conversational walk-through that works well on mobilePaperA print-style layout for PDFs and reportsTimelineStories about change over timeStoriesVertical tap-through cards for socialQuizLetting readers guess before you reveal the answerBentoA one-screen grid of key findings

You can set the accent, background and text colors and a chart palette so the result matches your outlet’s style.

A quick example

Say you have a public dataset of renewable energy share by European country from 2008 to 2022. (The figures below are illustrative.)

1. You drop the file in. Datastripes finds country, year and renewable_share and recognises the countries.

2. The draft opens with the leaders, then flags a turning point around 2015, when growth sped up.

3. You write your conclusion: “Countries that started early pulled further ahead after 2015.”

4. Datastripes shows the chapters behind it and says the data partly supports the claim, because two early starters stalled. You put that caveat in the piece.

5. You publish it as an article for the website and as Stories for Instagram, both in house colors.

Where it fits in your workflow

It’s useful for:

• A first look at a new dataset on deadline

• Checking and explaining a data-heavy press release

• Quick explainers, newsletter sections and social versions of a bigger investigation

• Small newsrooms and freelancers without a data desk

It won’t verify the source or methodology of your data, it can’t establish causation (two lines moving together are still just correlated), and it doesn’t replace talking to people.

Try it with the next spreadsheet on your desk

Start with a dataset you already know well, so you can judge whether the findings hold up. Drop it into datastripes.com, read the draft and see if it catches something you’d have missed.

If you’re a journalist, editor or newsletter writer, I’d like to hear what worked, what didn’t and which story you told with it first. Leave a comment or get in touch.