Common edits: text, color, and images

The three edits you'll make 90% of the time — how to do each one cleanly without breaking your layout.

Most A/B tests come down to three changes: rewriting some text, adjusting a color, or swapping an image. Here's how to do each properly in the visual editor.

Text edits

Click any text to select it. Either double-click to edit inline (great for short copy) or use the Edit text panel for longer copy with line breaks. The original text stays in the control variant — your edit only applies to this variant.

Tips:

  • Test specific copy changes (e.g. "Buy now" vs "Add to cart"), not whole paragraphs. The smaller the change, the clearer the signal.
  • If your site has dynamic text (product names, prices), don't edit the dynamic part — edit the labels around it.
  • For multi-language sites, edit only the locale that's running the experiment. URL targeting lets you scope per-language.

Color edits

Select an element and use the Color tab. You can change:

  • Background — the element's fill color
  • Text — the foreground color
  • Border — outline color

Stick to colors that already exist in your brand palette. The editor saves your recently-used colors so you don't have to copy-paste hex codes every time.

Image swaps

Click any image to select it. Choose Replace image. You can either upload a new file or paste a URL. We host uploaded images on our CDN with public-read access — no other configuration needed.

Recommendations:

  • Match the dimensions of the original — otherwise the surrounding layout shifts and your variant becomes a layout test, not an image test.
  • Optimize the file before uploading. PNG/JPG under 200KB. Use WebP if your audience supports it.
  • Don't break alt text. Whatever the original alt was, keep it (or improve it) on the variant for accessibility.

Combining edits

You can stack multiple edits in a single variant — change the headline, swap the image, and recolor the button all at once. Just remember: the more you change in one variant, the harder it is to know which change drove the result. If you want a clear answer, isolate one change at a time.

Was this article helpful?