Guide

AutoVariant Guide

Everything you need to create your first generated design.

  1. 1 Select a frame — AutoVariant uses it as the template every output is generated from.
  2. 2 Upload a CSV — one row per output. Column headers map to layer names in your frame.
  3. 3 Configure output — set naming structure and canvas layout.
  4. 4 Generate — AutoVariant creates one frame per CSV row, automatically.

Before you start

AutoVariant works by matching CSV column headers to layer names inside your selected frame. When the names match, AutoVariant maps the fields automatically.

Figma layer names
Title
Subtitle
Price
Product Image
→
CSV column headers
Title
Subtitle
Price
Product Image
01

Select a Frame

Select any frame on your Figma canvas. AutoVariant reads the text and image layers inside it — their names become the fields available for population from your CSV. The selected frame acts as the template: AutoVariant duplicates it once per CSV row.

Your selected frame
■ Frame
├─ Title
├─ Subtitle
├─ Price
└─ Product Image
⚠️
Avoid duplicate layer names within the same design. AutoVariant uses layer names to identify fields, so duplicate names can make mapping ambiguous and harder to manage.
💡
Tip: Use clear, descriptive layer names such as Title, Price, Product Image, or CTA Button. Any name will work, but consistent naming makes templates easier to maintain and understand.
02

Upload Your CSV

Upload a CSV file. Row 1 is your column headers — these correspond to the layer names in your selected frame. When names match, AutoVariant maps the fields automatically. Each subsequent row becomes one generated output.

Title Subtitle Price
Hoodie New Collection £49
Tote Bag Summer Drop £29
Cap Limited Edition £24

3 rows = 3 generated output frames. Column headers match the layer names from Step 01.

💡
Tip: Use ChatGPT to generate a CSV instantly — give it your column names and row count and it produces a clean, importable file in seconds.
03

Configure Output

Before generating, configure how AutoVariant names and arranges your outputs. Both settings are optional, but they become incredibly powerful when generating dozens or hundreds of variants.

Naming Structure

Control how generated frames are named. Combine static text with CSV values to create organised, export-friendly file names automatically. Example output names:

  • Product-Hoodie-Red
  • Product-Tote-Bag-Black
  • Product-Cap-Blue
⚡
Pro tip: Use the Name=Value format when generating component variants. This mirrors Figma's variant naming structure and can dramatically speed up component creation workflows.

Layout

Set how many variants appear per row on the canvas and the gap between them.

⚡
Pro tip: AutoVariant can populate multiple layouts at once. Place different format variations — square, portrait, landscape, banner, ad — inside the same selected frame. A single CSV run generates a complete asset suite across every format simultaneously.
04

Generate Variants

Click Generate. AutoVariant duplicates the selected frame once per CSV row, populates each copy with that row's data, and places all frames on the canvas. Once complete, every output is a normal Figma frame — edit, export, or hand off as usual.

✅
Test with a few rows first. Generate 5 rows before running a large batch. A mistake in the selected frame multiplies across every output — catching it early costs nothing.

06.Tips for best results

  • Use a single selected frame. Keep all variable layers inside one complete frame. Avoid components or nested frames with duplicate layer names — the plugin needs unambiguous targets to map reliably.
  • Large CSVs take longer. Batches of hundreds of rows require more processing time. Generation speed depends on batch size and Figma file complexity.

07.Common issues

Layer not appearing

Check the layer is inside the selected frame, is visible, and is a supported text or image type.

Text not updating

Check that the correct layer has been selected for mapping. For component variants, AutoVariant may expect a specific property format such as Property=Value. Also verify that your CSV column contains data and that the intended layer is being targeted.

Image not updating

Check the CSV contains valid image data for that column and that the image layer was included during field mapping.

Generated names look wrong

Review the Naming Structure settings in Step 03 — check the selected fields and separator match the format you expected.

Nothing generates at all

Select your frame on the canvas before opening the plugin. The plugin needs the frame selected to know what to duplicate.

Component variant not changing

Verify the mapped layer is using the correct variant naming structure. Many component sets require values such as Property=Value to match the component's variant options.

08.What else can you build?

AutoVariant works for any repeatable design where the layout stays fixed and the content changes — player cards are just one example.

Social graphics Hundreds of on-brand posts with different copy, dates, or images — generated in one batch.
Display ads Every size and copy variation for a campaign, driven from a single spreadsheet.
Product images Product cards, e-commerce tiles, and catalogue pages across an entire range.
Event graphics Speaker cards, schedule slides, name badges — one design, populated for every session or attendee.
Marketing campaigns Personalised banners, email headers, and out-of-home assets driven by data.
Any repeatable design If the layout stays the same and the content changes, AutoVariant handles it.

Generate your first batch in minutes

Install AutoVariant free and generate your first 5 outputs — no account required.