Integrations
Forty-one pieces of evidence, from three systems. One had been quiet for six weeks.
A product's evidence is scattered across tools that were never built to talk to each other. Connections exist so a claim can be checked against what customers did, said and paid for.
A system that can't reach your evidence is checking a snapshot that went stale the day it was taken.
Steps 01–02
An integration is whether a claim can be checked
This page is short on purpose. A directory and one argument.
FlowDesk's reframe rested on 41 pieces of evidence, and not one of them was created in Shorter Loop. They came from a support platform, a folder of interview transcripts, and notes taken on sales calls.
That's the normal case. Nobody's evidence lives where their strategy lives, which makes a connection the difference between a claim you can check and a claim you have to trust.
So the useful thing to show is provenance: for a given claim, which systems it was checked against, how recently, and what's missing.
Every connector on this page exists to answer one question: how do you know?
Steps 03–04
A stale connection is worse than a missing one
A gap you can see is a task. A gap you can't see is a wrong answer with a confident face.
FlowDesk's CRM connection had been silent since June. Nothing broke visibly. Reports still ran, dashboards still rendered, and every churned account quietly weighed the same as every other.
The reframe survived it, because support conversations and interviews were enough to find the pattern. What didn't survive was the ability to argue about size — and "we're losing our larger accounts" is a different conversation from "we're losing accounts."
Which is why freshness belongs on the claim itself, next to the conclusion, where somebody will read it. If a source hasn't spoken in six weeks, the conclusions drawn from it should say so.
Silence from a connector looks exactly like agreement.
What connects
Four categories, and what each one is actually for
Twenty-plus pre-built connections, plus an API and webhooks. Grouped by the question they answer rather than by vendor.
Delivery and engineering
Two-way sync for epics, features and issues, with the hierarchy preserved and status changes flowing back. Engineers never leave the tracker they already use.
Support and feedback
Conversations arrive as feedback items ready to cluster and deduplicate. This is where 21 of FlowDesk's 41 pieces of evidence came from, and none of them had been read as a set before.
CRM and customer data
Sales-call notes and account value attached to the feedback. Without it every request weighs the same, which is the gap that cost FlowDesk six weeks of a weaker argument.
Existing product tools
Import feedback, roadmaps and history from the product tool you're on now, or sync alongside it while you decide. Migration help is included rather than sold.
Freshness on the claim
Every conclusion shows what it was checked against, and when
On the opportunity itself, next to the number somebody is about to make a decision with.
Fair questions
What people ask about integrations
- How does two-way sync with a development tool work?
Epics, features and issues map across with their hierarchy intact, and status changes flow in both directions. Intent stays here, day-to-day status stays there, and nobody is asked to maintain two records of the same work.
- Can we migrate from the product tool we're on now?
Yes — feedback, roadmaps and historical data import from Canny, Productboard, Aha! and Pendo, and you can run alongside for a while rather than cutting over. Migration assistance is included.
- What if a connection breaks quietly?
That's the failure mode worth designing against. Last-sync time sits on the source and on any claim drawn from it, so a stale channel is visible next to the conclusion rather than buried in settings. In the demo, the CRM is silent for 43 days without raising a single error.
- Our source isn't on the list.
Custom API and webhooks cover it, and bulk CSV or JSON import covers anything sitting in a spreadsheet. The most decisive source in the demo is fourteen interview transcripts uploaded from a folder.
- Do we need to connect everything before we start?
No. Connect the one system that would settle your loudest current argument. FlowDesk starts with support conversations alone, and that is enough to overturn the premise of the quarter.
When did your customer data last tell you something you didn't already believe?
Connect one source. The first thing a live connection usually does is contradict something everyone was sure about.
FlowDesk is the Shorter Loop demo workspace. The provenance panel, the sync ages and the stale-source flag is live in it — open it and check any number here.