Tutorials · · 7 min read · By Data to Video Editorial
From CSV to Viral Video: A Step-by-Step Workflow
The exact end-to-end workflow for turning a raw CSV file into a published vertical data video, including cleaning, design and export.
The fastest way to publish a data video is to start with a clean CSV and never leave the keyboard. This workflow is what we use internally to ship two to three videos per hour, and it's what we recommend to every creator who asks how to scale output without sacrificing quality.
Step 1 — Decide the story before you touch the data
Open a notes app and write the headline of the finished video in one sentence. Examples: 'India's GDP overtook the UK in 2022.' 'Bitcoin's market cap quadrupled in five years.' Without a one-line story, you'll waste time animating data that doesn't lead anywhere.
Step 2 — Source the rawest data you can find
Always grab the rawest version available. Pre-processed charts on Statista or news sites are often filtered or rounded, which destroys the precision you need for a clean animation. Prefer:
- World Bank Open Data CSV exports
- Wikipedia table → copy-paste into Google Sheets
- FRED API for US macroeconomic series
- Kaggle datasets for sports and entertainment metrics
Step 3 — Clean the CSV in five minutes
Most CSVs fail to import because of three preventable issues. Fix all three before opening the editor:
- Strip currency symbols, percent signs and thousands separators ($1,200,000 → 1200000)
- Convert merged or wide headers into a single header row
- Drop footnote markers ('1', '*') from numeric cells
Step 4 — Shape the data wide, not long
Almost every animation engine expects 'wide' format: one row per entity, one column per time step. If your CSV came in 'long' format (one row per entity-year combination), use a pivot in Google Sheets or Excel to reshape it before importing.
Step 5 — Pick the template that matches your story
The story you wrote in step 1 determines the template:
- A ranking story → bar chart race
- A reveal story → top 10 countdown
- A growth story → animated timeline
- A rivalry story → head-to-head comparison
Step 6 — Write the hook overlay
Type your one-line story into the hook field. Don't overthink it — the version you wrote in step 1 is almost always the right one. Resist the urge to soften it; confrontational hooks outperform polite ones by a wide margin.
Step 7 — Render at 1080×1920
Always export vertical. Horizontal exports cut TikTok and Reels reach roughly in half. The MP4 should be H.264, 30 fps, with stereo audio if you added a music bed.
Step 8 — Caption, post, repeat
Add a single-sentence caption that mirrors the hook. Hashtags are optional in 2026 — the algorithm reads the on-screen text, not the tags. Post immediately; data videos lose relevance fast.
Treat each video as one of fifty, not one of one. Speed and frequency beat polish every time in this format.
Frequently asked questions
- Does the CSV need a specific encoding?
- UTF-8 is safest. Excel sometimes saves as Windows-1252 by default, which breaks non-ASCII labels — re-save explicitly as UTF-8 CSV before importing.
- What if my data has gaps in the time series?
- Either interpolate explicitly or leave the cells blank. The animation engine treats blanks as 'no data' and skips that frame for the affected series.
- How big should the CSV be?
- Practically anything under 5 MB imports fast. The visual limit is around 10 rows × 30 columns — beyond that the chart becomes unreadable on mobile.
About the author
Data to Video Editorial — Editorial team. We test every format we write about. Our editorial team publishes data videos daily across niche TikTok and YouTube accounts and shares what actually moves the needle for creators.