Definition
Microcopy is the small, functional text in user interfaces — button labels, form hints, error messages, tooltips, empty states — that guides users through interactions one decision at a time.
Microcopy is where UX writing earns its keep. A well-worded error message can be the difference between abandonment and recovery. A button labelled 'Save and continue' converts at a different rate than one labelled 'Submit'. The headline gets all the attention; microcopy moves the metric.
Good microcopy is invisible — users don't notice it, they just succeed. Bad microcopy is loud — it confuses, frustrates, or makes users feel stupid. The discipline is to write so users complete the task, not to show off the writing.
Origin
The term was popularised by Joshua Porter in a 2009 blog post, though the practice is older — Apple's Human Interface Guidelines have included writing standards since the 1980s. The 2017 book 'Microcopy: The Complete Guide' by Kinneret Yifrah codified the practice.
How it works
- Inventory the microcopy in your product — buttons, errors, empty states, tooltips.
- Categorise by function: instructional, confirmational, error-recovery, encouragement.
- For each, ask: what does the user need to know to succeed at this exact moment?
- Test variants. Microcopy A/B tests are some of the highest-ROI tests you can run.
- Standardise: a tone-of-voice document and patterns library prevent drift.
When to use it
Use when
- On any interaction where the user makes a decision (forms, errors, empty states, CTAs).
- After usability testing reveals confusion at specific moments.
- When tone needs to shift across the product (transactional vs emotional moments).
Skip when
- Treating every UI string as 'microcopy'. The label on a settings field is microcopy; a marketing headline isn't.
- Without context. Microcopy that works on a free tier flops on enterprise.
Key metrics
- Form completion rate (label/hint changes can swing it 5–15%).
- Error recovery rate (% of users who succeed after an error).
- Click-through on CTA buttons (microcopy A/B tests).
- Empty-state engagement (% of users who do something after hitting empty state).
Examples
- Changing the button label from 'Submit' to 'Get my free quote' lifted clicks 32%.
- The error message used to say 'Invalid input.' Now it says 'Phone number needs the country code, e.g. +1 415...'. Recovery rate doubled.
- Microcopy is a competitive moat — it's hard to copy because it's small and specific.
In practice at Makreate
Makreate UX engagements include microcopy as a deliverable, not an afterthought. A recent SaaS client had a perfectly designed signup form with 27% completion. We rewrote the field labels, hint text, and error messages — no design changes — and completion climbed to 41% in three weeks. Most of the lift came from one change: replacing 'Required' on every field with specific, helpful hints about what to enter.
UX Design →Common mistakes
- Letting engineers write microcopy. They optimise for accuracy, not user success.
- Writing once and never testing. Microcopy A/B tests are cheap and high-ROI.
- Inconsistent tone — formal in onboarding, casual in errors, marketing-speak in nav.
- Generic error messages. 'Something went wrong' is a sin. Tell the user what to do next.
Frequently asked
Who should write microcopy?
A UX writer if you have one; a designer or copywriter if you don't. Engineers should not write microcopy without review.
How do I test microcopy?
Same way you test any UI — A/B testing for high-traffic surfaces, usability testing for moments of friction, qualitative review for tone consistency.
Microcopy vs UX writing?
Microcopy is the smallest, most functional UX writing. UX writing covers everything from microcopy up to onboarding emails.