# Component Analytics: Measure Which Design System Parts Teams Really Use

> Learn what to measure, where the usage numbers come from, what they cannot tell you, and how to turn them into decisions about your components.

**In short:** Component analytics measures which design system parts teams really use, where the numbers come from, what they cannot tell you, and how to turn them into decisions.

From "Design systems", a free course by Venkat R. Murty: https://designervenkat.online/learn/design-systems/component-analytics

## What You Will Learn

Design systems are built to be reused. Their parts, called components, are meant to appear again and again in real products. But building a part is only half the job. Someone also has to check whether people pick it up, change it, or ignore it. That checking is called component analytics.

This lesson covers three things. First, the numbers that matter most. Second, where those numbers come from and what they cannot tell you. Third, how a team turns a chart into a decision. A worked example walks through a real Figma screen step by step, and a short quiz checks your understanding.

- **Component:** A reusable design part, such as a button, a dropdown, or a card.
- **Instance:** One copy of a component placed on a screen or inside a file.
- **Library:** The shared file or package that holds a team's published components.
- **Adoption:** How many teams, products, or files use the system's parts.
- **Detach:** Cutting one copy loose from its shared original, so it no longer follows changes to the library.
- **Override:** Changing a copy's look or content while it stays linked to the original.
- **Coverage:** The share of a screen or product built from system parts instead of custom work.
- **Deprecate:** Marking a part as old, so teams stop using it and it can be removed later.
- **Code scan:** An automatic check of product code that finds where shared parts appear.

> **Tip:** Write your question before you open any dashboard. A question such as 'Do designers change the Dropdown more than other parts?' tells you which numbers to read.

## What does component analytics mean?

Component analytics is the practice of collecting and reading data on how a design system's components are used across products. The data can come from design files, from product code, or from both. Teams do not collect it to admire a chart. They collect it to learn which parts help people, which parts get in the way, and which parts nobody needs.

The data answers three plain questions. Where is a part used? How often do people change it after placing it? How much of the product is built from shared parts at all? The answers help a team decide what to improve, what to promote more strongly, and what to retire.

## What should you measure?

Three measures do most of the work. Adoption shows where parts are used. Detachments and overrides show where parts fail to fit real needs. Coverage shows how much of the whole product the system reaches. Start with these three before you add anything else.

### How do you measure adoption?

Adoption counts how many instances of a component exist, and in how many places. A common way to read it is instances per product or per team: how many copies of a button sit in each team's files, or how many products import it. Figma's built-in analytics report component instances and insertions by team, so a design lead can see which groups rely on a part and which ignore it.

> **Tip:** A high count is not proof of value on its own. A design-file count shows what a designer chose to place, not what reached customers. Pair it with a check of real product code.

Connected is not the same as used. In one practitioner's account, a dashboard showed that most of the company had connected to the shared library, yet the share of components truly reused was far lower. Count the people who use a part, not only the people who installed the library.

### How do you measure detachments and overrides?

A detach happens when someone cuts a copy loose from the shared original. After that, the copy stops following library updates. An override is a change made to a copy that stays linked, such as a different label or color. Both are useful signals, because frequent detaching often means a part does not meet real needs. Either it has gaps, or it is too rigid for what teams build.

Detach counts do not explain themselves. Figma's library analytics show how often a component was detached and by whom. A community request on Figma's forum points out that the tool does not capture the reason behind each detach, such as a missing variant or a bug.

High detachment is not always bad. Some parts are meant to be copied, such as a screen template that teams use as a starting point. Figma's guidance says a high detach rate for a part like that is not necessarily a bad sign.

### How do you measure coverage?

Coverage is the share of a product's interface built from system components compared with custom work. It is usually shown as a percentage. One guide describes it as the percentage of user interfaces built with system components versus custom-built ones.

Teams measure coverage in different ways. Some count imports in code. Some look at the screens. Some count the rendered parts that come from the system. Use one method consistently, so that numbers from one month can be compared with the next.

Coverage can also apply to tokens, which are named design values such as a brand color or a standard spacing step. Tracking tokens and components together gives a fuller picture of how the system is used.

Some guides offer target numbers. Treat that as one author's rule of thumb, not a standard, and set your own target from your starting point.

## Where the Numbers Come From

There are two main places to collect data. Design files show what designers place and change. Product code shows what actually ships to customers. Each source is partial, so the clearest picture comes from reading both together.

### Figma Library Analytics for Design Files

Figma's library analytics are available on its Organization and Enterprise plans. They show how published libraries, components, styles, and variables are used across a company. The Analytics tab lists each component's total instances, inserts, and detaches over the last 30 days.

You can sort the component list by inserts, total instances, or detachments. That makes the most used and least used parts easy to spot. Figma also offers an expanded Library Analytics API for Enterprise customers, which reports insertion and detachment events for teams that want to build their own reports.

Figma updates the data daily and keeps up to a year of history, so you can compare this quarter with last year.

- **Some use may not show:** The numbers may not include every file where a component is used, such as other people's private drafts. Treat them as a guide, not a full count.
- **Copied files can skew counts:** Duplicating a whole file can add many insertions at once, so check a sudden jump before reading it as real adoption.
- **Copies can inflate counts:** Read insert counts with care when detaches are high.
- **Trends can mislead:** Releases and maintenance work can push detaches up, and quiet months can push them down. A single month rarely tells the whole story.

### Code Scanning and Usage Tracking for Production

For product code, teams count how often system components are imported and rendered. Some scan the codebase with a tool. For example, React Scanner is a tool that reads React code to pull out which components are used and which settings they receive. Other tools collect usage statistics from many repositories and turn them into adoption reports.

- **Installed is not used:** A package download or a library connection shows that someone set something up. It does not show that the parts appear on screens. Check the screens themselves.
- **Methods differ:** Coverage can be measured through code imports, visual checks, or render counts. Two teams using different methods will get different numbers, so compare like with like.
- **Code shows the gap, not the cause:** A scan can show that a team built its own card. It cannot show why. For that you need to ask the people involved.

> **Tip:** Read design data and code data side by side. A part that designers place often but developers rarely import is a clue worth chasing.

## Worked Example: The Dropdown That Keeps Getting Detached

The team in this example is made up, but the screens it uses are the same kind Figma shows in its library analytics. Follow the steps the team takes once a month.

1. **Open the analytics:** From the file browser, the team opens Libraries under All teams or All workspaces, then selects the library to see its analytics. The panel has two tabs: one for high-level numbers and one for detailed asset metrics.
2. **Filter to components and sort:** They set the type filter to components, then sort the list by detachments. The list now puts the most detached parts at the top.
3. **Read the top rows:** The Button sits near the top with thousands of instances and inserts, so it is clearly well used. The Dropdown shows a high number of detaches across most files. Those two rows tell the first part of the story.
4. **Check the context:** Before drawing a conclusion, the team asks whether the Dropdown was ever meant to be copied. A detach is not automatically a failure, so they check the part's purpose first.
5. **Ask the people who detach:** The team interviews designers from several groups and asks what they changed and why. A short interview is the fastest way to learn whether a gap, a bug, or a missing feature is behind the change.
6. **Find the gap:** The interviews reveal that the Dropdown has no searchable variant. Designers who need search build their own search field each time, and that custom work causes the detaches.
7. **Act, then measure again:** The team adds a searchable variant to the next release. In the following months, they check whether the Dropdown's detach count falls and whether more teams use it.

## Turning Numbers Into Decisions

A number alone does not tell you what to do. Use a simple loop: spot a signal, ask why, choose an action, then measure again. Figma's own case study describes a similar loop: examine the component for obvious problems, reach out to the teams involved, and then create a fix as a ticket in Figma or in code.

### Spot the Signal and Ask Why

1. **Spot the signal:** Look for parts with very low use, high detach counts, or low coverage. Figma's guidance suggests watching which parts see the most use and which see little or none.
2. **Ask why:** Talk to the people involved, read the part's documentation, and check whether the part is hard to find. The answer tells you which fix you need.
3. **Choose an action:** Pick one of four paths: improve, promote, fix the documentation, or retire. The next subsection explains each one.
4. **Record and review:** Add the finding to your list of planned work, with an owner and a date to check the numbers again.

### Improve, Promote, Document, or Retire

- **Improve:** A heavily detached part may need a new variant, a fix, or more flexibility. Work out which one before you change anything. A part that is detached often is not in the right form for its job.
- **Promote:** A part that teams use well and that solves a common problem deserves more attention. Give it better examples, clearer guidance, and a visible place in the library.
- **Fix the documentation:** Sometimes the part is fine, but teams cannot find it. A high token use with low component coverage can mean that every team builds the parts again, because the library is hard to search or the parts are not documented.
- **Retire:** If a part is rarely used and no longer worth maintaining, mark it as deprecated and plan its removal. Microsoft's designers have said that they look for components no one uses and that are not worth keeping.

> **Tip:** Look at trends, not single snapshots. A spike can come from a release, and a quiet month can come from a holiday. Compare several months before you act.

## Common Mistakes

- **Counting insertions as adoption:** Design-file counts show intent. They do not prove that the product uses the part. Always pair them with a code-phase or product check.
- **Judging a part by one number:** A detach ratio can flag a part worth a closer look, but it cannot say whether the cause is a user problem, internal maintenance, or a seasonal change.
- **Reporting numbers without decisions:** A dashboard that looks impressive does not help the system. Each number should lead to a question, a conversation, and a clear next step.

## Check Your Understanding

Try these short scenarios. Each one asks you to use the ideas from this lesson.

**Quick check:** Your team's card component appears on many screens, yet it is used on only 10% of the screens where cards appear. Most teams build their own version. What should you check first, and what might you do next?

A. Only the total number of card inserts in Figma, then add a banner to every page asking people to use it.
B. Adoption by team, detach counts, and code coverage for cards. Then interview the teams that built their own to learn why, and decide whether to improve, document, or retire the card.
C. Nothing. A 10% share is normal, so coverage only matters for buttons and text fields.
D. Delete the card component at once, because teams clearly do not want it.

**Answer:** B. Several measures show where the gap is, and talking to teams shows why it exists. Only then is a fix or a retirement a sensible choice.

**Quick check:** A banner component has few inserts across all teams. One team detaches it in almost every file, and no other team uses it. What is the best first step?

A. Mark the banner as deprecated across the whole system today.
B. Count the detaches as proof the banner is popular, and celebrate the number.
C. Talk with that one team to learn why they detach it, and check whether a missing variant or a bug explains the change.
D. Remove that team's access to the library so the detaches stop.

**Answer:** C. Frequent detaches point to a need or a problem. A conversation finds the cause before anyone decides whether to change or retire the part.

**Quick check:** A dashboard shows that 95% of product teams have connected to the shared library. Developers still say they rarely use its parts. Which measure should you add to see the real picture?

A. The number of people who opened the library's documentation page.
B. Coverage: the share of each product's screens built from system parts compared with custom work.
C. The number of teams who joined the library's chat channel.
D. The count of library installs in the last week.

**Answer:** B. A connection shows installation, not use. Coverage compares shared parts with custom work, so it tests real use in the product.

## Key Takeaways

Component analytics turns usage data into design decisions. Measure adoption, detaches and overrides, and coverage. Get the numbers from Figma's library analytics for design files and from code checks for production. Know the limits of each source. When a number looks wrong or surprising, ask the people involved, then improve, document, promote, or retire the part and measure again.

> **Tip:** Measure early and keep measuring. Figma's guidance is to set goals first, so you know which numbers matter. Other lessons in this course cover measuring the design system's tools and its overall health.

## Sources

- [Figma blog: Design system 104, Making metrics matter](https://www.figma.com/blog/design-systems-104-making-metrics-matter/)
- [Figma Help Center: View and explore library analytics](https://help.figma.com/hc/en-us/articles/360039238353-View-and-explore-library-analytics)
- [Figma Help Center: Library analytics (who can use it and what it shows)](https://help.figma.com/hc/en-us/articles/360039238353)
- [Figma Developers: Library Analytics REST API endpoints](https://developers.figma.com/docs/rest-api/library-analytics-endpoints/)
- [Figma best practices: Five ways to make the most of your Design System Analytics](https://www.figma.com/best-practices/make-the-most-of-design-system-analytics/)
- [Figma blog: How Pinterest's design systems team measures adoption](https://www.figma.com/blog/how-pinterests-design-systems-team-measures-adoption/)
- [Medium: Measuring Design System Adoption in Figma (tools and practices)](https://davidveraco.medium.com/measuring-design-system-adoption-in-figma-tools-comparisons-and-best-practices-c1783fc2c06b)
- [Adora: Design system analytics, measure component usage and impact](https://www.adora.ai/blog/design-system-analytics/)
- [DNSK.WORK: Design System Adoption Metrics Lie](https://dnsk.work/blog/design-system-adoption-metrics-lie-measuring-installation-measuring-value/)
- [Mews Developers: Building a design system adoption metric from production data](https://developers.mews.com/design-system-adoption-metric-building/)
- [Mironsoft: Measure design system adoption instead of guessing](https://www.mironsoft.de/en/blog/design-system-adoption-audit-metrics)
- [CabinCo: Design System Best Practices That Drive Adoption](https://cabinco.com/design-system-best-practices-drive-adoption/)
- [Netguru: How to Master Design System Metrics](https://www.netguru.com/blog/design-system-metrics)
- [Waysconf: Design system analytics, measuring the adoption of Figma components](https://www.waysconf.com/lecture/design-system-analytics-how-to-measure-the-adoption-and-success-of-figma-components)
- [Medium: Are your Figma Design System analytics high as a kite?](https://iknowdavehouse.medium.com/are-your-figma-design-system-analytics-high-as-a-kite-c5b1440a112b)
- [Figma Community forum: feature request on detach reasons](https://forum.figma.com/suggest-a-feature-11/trigger-plugin-widget-hooks-on-component-instance-detach-51297)
- [UX Design (Medium): The challenge of Design Systems adoption metrics](https://uxdesign.cc/design-systems-adoption-metrics-over-the-past-5-years-b389308d6663?gi=aada7a13f375)
- [Medium: Design Systems metrics, lessons from leading a design system at an edtech company](https://medium.com/@andrerolla/design-systems-metrics-what-ive-learned-leading-the-ds-of-the-biggest-edtech-in-latin-america-f9f4fffc06c2)
