How-To Guide12 min read
How to Link Customer Feedback Themes to Linear Issues and Projects
Last updated: · By Matt Teixeira, Founder and CEO
On this page
To go from customer insight to shipped work, don't link a feedback theme straight to a Linear ticket. Size the theme into one problem, link that problem to one Linear project with its evidence attached (unique accounts, ARR, three quotes, a success metric), and log each customer's specific ask on the issue that solves it with Linear Customer Requests. Let Linear own status and your feedback tool own the evidence, then recheck the theme 14 and 30 days after release to see whether complaints fell.
Key Takeaways
- Link a sized problem to a Linear project, and one customer's specific ask to the issue that solves it.
- Write the evidence brief and success metric before you open Linear; the handoff is where the "why" gets lost.
- Give every field one owner: Linear for status, assignee, and dates; the feedback tool for evidence and the success metric.
- Linear Customer Requests is on every plan; automatic intake from Intercom and Zendesk needs Business, at $16 per user per month billed yearly (Linear pricing).
- After ship, count unique accounts raising the theme in equal windows before and after, and post the verdict as a Linear project update.
Why Does Customer Evidence Get Lost Between the Theme and the Ticket?
Because they come in different sizes. A theme is a topic dozens of customers raise over months; a Linear issue is work one engineer finishes in days. Link them directly and you either paste 40 quotes into a ticket nobody reads or scatter single quotes across 12 tickets and lose the count.
The fix is a middle layer: a sized problem, which most feedback tools call an opportunity. The opportunity, not the theme, becomes the Linear project.
| In your feedback tool | In Linear | What to attach |
|---|---|---|
| Theme, a topic such as "Data export" | Nothing directly, or a Linear initiative for a quarter-long bet | A link to the theme's live view |
| Opportunity, a sized problem | One Linear project | The evidence brief: accounts, ARR, quotes, success metric |
| One customer's request or bug | A customer request on the issue that solves it | The account, the quote, and the Important flag if it blocks them |
Rule of thumb: one opportunity, one Linear project. An opportunity that needs two projects is probably two problems. A project that serves three opportunities needs one named as primary, so the after-ship check has a single target.
Step 1: Turn the Theme into a Sized Opportunity
A theme tells you where customers are talking; an opportunity says what is wrong and how much it matters. Write the problem as one sentence in customers' words, then size it on four numbers: unique accounts in the last 90 days, their combined ARR, severity, and trend.
Set a bar before anything reaches Linear, such as five or more accounts in 90 days, or one account that cited the problem in a churn or lost deal. Below the bar, opportunities stay in your feedback tool and keep collecting evidence. To rank the rest, see how to decide what to build next.
Step 2: Write the Evidence Brief Before You Open Linear
Write it in the format engineers will see at the top of the Linear project. Budget 15 to 20 minutes per opportunity if your evidence is already in one place.
## Problem
[One or two sentences in the customers' words]
## Evidence (as of [date])
- [n] accounts in the last 90 days · $[ARR] · trend: [rising / flat / falling]
- Severity: [blocks a core job / costly workaround / annoyance]
- "[Quote]" ([account], [source link])
- "[Quote]" ([account], [source link])
- "[Quote]" ([account], [source link])
- Full evidence: [link to the opportunity or theme]
## Success metric
[What should change in customer feedback] · check on [ship + 14 days] and [ship + 30 days]
## Out of scope
[Related requests this project will not solve]
Pick three quotes, not thirty: one that shows the job, one that shows the cost, and one from your most important segment. Engineers read three and skim thirty. If you can't write the success metric, the project is not ready for Linear.
Step 3: Create or Link One Linear Project per Opportunity
Search Linear first. A second project for the same problem splits the customer count and the after-ship signal.
- Name the outcome, not the theme. "Large exports finish for admins" beats "Export improvements."
- Use both text fields. The problem sentence goes in the short description, which shows in list views; the full brief goes in the project overview.
- Link back to the evidence. Put the link in the overview so an engineer reaches the source quotes in one click.
- Split the work into issues. Give each issue only the quotes that apply to it.
Step 4: Attach the Customers Who Asked with Linear Customer Requests
Customer Requests ties customers to the work they asked for. Customers are organizations with fields such as revenue, size, tier, status, and owner, and requests attach to issues and projects (Linear docs). A project's Customers tab rolls up requests from its issues, and you can sort projects by customer count and revenue (Linear changelog).
- From support: Intercom, Zendesk, and Front create requests on Business and Enterprise. Linear Agent can turn an Intercom or Zendesk conversation into triaged issues linked to the customer in one click (Linear changelog).
- From sales: Gong creates requests from call transcripts on Enterprise, and Salesforce creates issues from cases as an Enterprise add-on (Linear Salesforce docs).
- From everywhere else: by hand, through the GraphQL API, from Slack, or from tools such as Plain, Pylon, and Attio (Linear integrations).
Two habits keep the data useful. Flag a request Important only when it blocks the customer, so the flag keeps its meaning, and sync revenue and tier from your CRM or support tool instead of typing them.
Linear Agent, on all plans, can also "extract common requirements across customer requests" to start a spec (Linear changelog). That covers requests already in Linear; themes across calls, interviews, and surveys still come from Step 1.
Step 5: Decide Which Tool Owns Each Field
"Two-way sync" means something different in every tool, so decide ownership before you connect anything. When two tools can both edit status, they eventually disagree and nobody knows which is right.
| Field | Owner | Shows up in the other tool as |
|---|---|---|
| Status, assignee, cycle, target date | Linear | Read-only fields in the feedback tool |
| Problem, evidence, success metric | Feedback tool | A link, or an overview written once at handoff |
| Who asked | Pick one: Linear Customer Requests or the feedback tool | A count with a link |
| Post-ship feedback and verdict | Feedback tool | A Linear project update |
Link both ways: one click from the Linear project to the evidence, and one from the opportunity to the Linear project. If you get only one direction, choose Linear to evidence. Anyone can find status in Linear; the evidence is what goes missing.
Step 6: Check Impact After the Project Ships
When the project completes, book two checks, 14 and 30 days after release. At each, count unique accounts raising the same theme in equal windows before and after, and sort the new quotes into "same problem," "new problem," or "unrelated."
Call it fewer, about the same, more, or too early to tell, and post the verdict as a Linear project update. The full method is in how to measure whether a release reduced customer complaints.
Tell customers when the work ships, but tell your team it worked only after the 30-day check. Linear can notify you when a customer's requested issues or projects ship if you subscribe from that customer's page (Linear changelog). For what to send, see how to close the customer feedback loop.
Worked Example: From a Feedback Theme to a Shipped Linear Project
This example is illustrative: a B2B analytics company with about 300 accounts.
- Theme: "Data export" has 140 mentions in 90 days, which is too broad to build.
- Opportunity: "CSV exports over 100,000 rows time out, so admins rebuild reports by hand." It covers 23 accounts and $540k ARR, it is rising, and two churn notes name it.
- Success metric: 11 accounts raised export timeouts in the 30 days before release; the target is three or fewer in the 30 days after.
- Linear: the project "Large exports finish for admins" gets the brief in its overview and four issues: a background export job, a ready email, a progress bar, and retries. Fourteen Intercom and Zendesk requests land on the right issues, and three are flagged Important.
- Day 14: four accounts raise exports in the two weeks after release, against five in the two weeks before. Two of the four say the emailed download link expires too fast, a new problem that becomes a new issue.
- Day 30: two accounts still report timeouts, against 11 before, which beats the target. "Fewer" goes into a Linear project update; the requesting customers already heard at launch.
Day 14 is the lesson. The count looked flat, but the quotes showed the original problem mostly gone and a smaller one in its place. Count accounts, then read what they said.
Common Mistakes to Avoid
- Counting request rows instead of accounts. One customer's request copied onto five issues is still one account.
- Letting the brief go stale. If a project waits more than a month to start, refresh the counts and quotes before kickoff.
- Linking once and never again. When a Linear project is split or merged, move the evidence link, or the after-ship check measures the wrong work.
- Leaving evidence-free work unlabeled. Security, infrastructure, and strategic bets often have no customer requests. Label them "no customer evidence" so the gap is a choice, not an oversight.
Tools That Help
Linear plus a spreadsheet can run this whole process: one row per opportunity with account counts, the Linear project link, and the two check dates. Dedicated tools differ mostly in what they create in Linear and what they read back. To analyze Linear's own data as a feedback source instead, see how to use Linear customer requests with customer feedback.
| Tool | What links to Linear | Creates in Linear | Reads back from Linear |
|---|---|---|---|
| Deck (source) | Deck Projects and Initiatives, to Linear projects and initiatives | A Linear project, with an optional AI-written overview | Title, description, status, owner, and dates, daily |
| Productboard (source) | Specs and features, through the Spark Linear connector | Issues, and a project split into issues from an approved spec | Progress on request; Spark suggests status updates |
| Dovetail (source) | Highlights, docs, ideas, and data points | Issues or projects, or a link to existing ones | Not described in its docs |
| Canny (source) | One idea to one issue or project | Issues | Status, synced both ways for Ideas |
| Cycle (source) | Feedback and docs | Issues or projects from highlighted text | Status of related issues |
- Canny: for teams whose main input is a voting board. The integration is on Pro and Business, with Pro from $79 per month billed yearly (Canny pricing), and merging ideas drops the merged idea's Linear links (Canny Help).
- Dovetail: for research teams sending interview highlights to engineering. Each user connects their own Linear account (Dovetail docs), and plans are Free and custom Enterprise, though neither page says which plan includes Linear (Dovetail pricing).
- Productboard: for teams that write specs in Productboard. Linear is not among its native delivery integrations, which include Jira and Azure DevOps; it connects only through each user's Spark connector (Productboard integrations).
- Cycle: for teams that write specs next to the feedback behind them. Linear's listing says Cycle can "automate status updates from one platform to the other" but not in which direction (Linear integrations), so test it first.
Where Deck Fits
Deck covers Steps 1, 2, and 6 and connects to Linear at the project level. It groups feedback into Themes, ranks problems in the Opportunity Backlog, and turns the one you pick into a Deck Project that links back to the opportunity, starts with its insights, and can draft the Overview, Evidence, and Success Metrics with citations to those insights (Deck Projects docs).
- Into Linear: create a Linear project with a short description (up to 255 characters) and an optional AI-written overview, link an existing one, or import a Linear initiative with its child projects (Deck Linear docs). The opportunity-to-delivery playbook and Linear roadmap import playbook walk through both.
- Field ownership: Deck refreshes title, description, owner, dates, and status from Linear daily and never overwrites the Project's Overview, Evidence, or Success Metrics.
- After ship: new high-confidence insights from Slack, support, NPS, Gong calls, and other sources keep attaching to the Project. Coding agents can read the Evidence and Success Metrics through Deck's MCP server.
- Pricing: Free is $0 with one integration, and Linear does not count toward it. Starter is $99 a month, or $79 billed annually (Deck pricing).
- Limitations: Deck links Projects to Linear projects, not to individual issues, and it does not create Linear customer requests. Links are one to one: each Deck Project has one Linear project, and each Linear project has one Deck Project. Deck writes to Linear only when it creates a Linear project; after that, updates come from Linear once a day. Child relationships do not sync after import, and only org admins can connect Linear.
Frequently Asked Questions
Sources
- Linear Docs: Customer Requests
- Linear changelog: Customer requests on projects
- Linear changelog: Linear agent for Intercom, Zendesk, and Gong
- Linear changelog: Introducing Linear Agent
- Linear Docs: Salesforce
- Linear pricing
- Linear integrations: Customer experience
- Linear integrations: Cycle
- Canny Help: Linear integration
- Canny pricing
- Dovetail docs: Linear integration
- Dovetail pricing
- Productboard: Linear MCP connector
- Productboard integrations
- Deck docs: Linear integration
- Deck docs: Projects
- Deck pricing
Sources checked November 2, 2026.