A Zendesk integration connects Zendesk to another tool so data and actions move between them without copy-paste: a CRM record in the ticket sidebar, a Jira issue created from a ticket, an order pulled from Shopify. Jump straight to the best Zendesk integrations by job or the five setup steps.
Zendesk handles the core job well. Every request becomes a trackable ticket, and email, chat, voice and social support channels land in one thread. What it does not hold is the context around the ticket: plan, orders, open bugs, past fixes. Integrations bring that context in, and this guide covers which apps to pick, how to connect them and which method fits each job.
TL;DR
- Start from the gap. Name the context your agents keep leaving Zendesk to find, then pick one integration that closes it.
- Marketplace app first. Most popular tools (Salesforce, Slack, Jira, Shopify) have an app you install from Admin Center in minutes.
- No-code tools for glue. Zapier-style platforms connect tools that have no app, using triggers and actions.
- API or ZIS for custom logic. Build on the Zendesk REST API or its integration services platform (ZIS) when the flow is two-way or multi-step.
- AI on top. An AI layer such as Pluno uses the context your integrations provide to resolve complex tickets, not just display data.
- Keep it small. One to three integrations, one clean customer identifier, then add more.
How to set up a Zendesk integration in 5 steps
The steps below cover the common case, a Marketplace app. You need the Zendesk admin role. Tool-specific guides in the table below have the click-level detail for each app.
- Name the data gap. Write down the one piece of context agents keep tab-switching for, such as order status, plan tier or a linked bug. That decides the tool, the direction of sync and the fields you need.
- Find the app in the Marketplace. Search the Zendesk Marketplace for the tool and check the listing for who built it, what it syncs, the plans it needs and its price. Read the How To Install tab for any pre-install tasks. First-party integrations, such as Salesforce, Slack and Shopify, are set up in Admin Center under Apps and integrations > Integrations > Integrations instead.
- Install and authorize. Click Install in the upper-right of the listing, configure the app settings, then click Install again. Most apps then ask you to sign in to the other tool with an admin account to grant access. Installed apps appear in Admin Center under Apps and integrations > Apps > Zendesk Support apps.
- Map the fields and test on one ticket. Match identifiers first (email, external ID or account ID), then the fields you want in the sidebar. Open a test ticket, check the right record appears, and trigger one write action if the app has one.
- Watch it for two weeks. Check failed syncs, duplicate users and agent feedback. Give the integration a named owner, and note where its settings live so the next admin can find them.
Tip: if the tool has no Marketplace app, the same five steps apply to a no-code connector or an API build. Only step 3 changes: authorize with OAuth. Zendesk is retiring API tokens: no new tokens can be created from 27 October 2026, and all of them stop working on 30 April 2027.
Best Zendesk integrations by job
Already know which tool you want to connect? Each guide below covers what the Zendesk integration syncs, how to set it up and where it falls short. The last four rows link to the vendor's own page.
| Job | Tool | What the integration covers | Guide |
|---|---|---|---|
| CRM | Salesforce | Accounts, contacts, opportunities and tickets in one view for sales and support | Zendesk Salesforce integration |
| CRM | HubSpot | The two official apps and what each one syncs | Zendesk HubSpot integration |
| CRM | Pipedrive | Deal and contact data for the sales-support handoff | Zendesk Pipedrive integration |
| Marketing | ActiveCampaign | Support tickets in your CRM and marketing automation | Zendesk ActiveCampaign integration |
| Team chat | Slack | Ticket alerts and working tickets from a channel | Zendesk Slack integration |
| Team chat | Microsoft Teams | Ticket notifications and collaboration inside Teams | Zendesk Microsoft Teams integration |
| Engineering | Jira | Escalating tickets to engineering and syncing status back | Zendesk Jira integration |
| Projects | Asana | Turning tickets into project tasks without copy-paste | Zendesk Asana integration |
| Gmail | The native connector, SMTP relay and no-code options | Zendesk Gmail integration | |
| Billing | Stripe | Payments, invoices, disputes and subscriptions inside a ticket | Zendesk Stripe integration |
| ERP | NetSuite | Prebuilt apps, iPaaS and custom builds for ERP data | Zendesk NetSuite integration |
| Website | Webflow | Live chat and form submissions routed into tickets | Webflow Zendesk integration |
| E-commerce | Shopify | Shopify order details in the ticket, plus refunds and cancellations from Zendesk | Zendesk's Shopify setup article |
| Projects | Trello or ClickUp | ClickUp tasks created from a ticket and linked back to it; Zendesk tickets and views attached to Trello cards | ClickUp's Zendesk article, Trello's Zendesk article |
| No-code automation | Zapier-style tools, such as Zapier, Make or Workato | Ticket triggers and create-or-update actions that link Zendesk to thousands of other apps | Zapier's Zendesk page |
| AI assistants | ChatGPT, Claude and other MCP tools | Zendesk's MCP client lets AI action flows call tools on outside MCP servers | Zendesk's MCP client announcement |
Marketplace app, no-code tool or API: which to use
There are five ways to build a Zendesk integration, and most teams mix them. Pick by how deep the sync goes, which way data flows and who will maintain it.
| Method | Setup effort | Direction | Who maintains it | Cost model | Best for |
|---|---|---|---|---|---|
| Marketplace app | Low, admin UI | Often one-way into the sidebar | The app's developer | Free or a monthly subscription per app | Showing another tool's data in the ticket |
| Native connector | Low to medium | Varies by object | Zendesk | Included, but some need specific plans | Salesforce, Slack and other first-party links |
| No-code tool, such as Zapier or Make | Low to medium | Trigger and action, either way | You, inside the tool | Per task or run; Zendesk is a Premium app on Zapier, so it needs a paid plan | Glue between tools that have no app |
| REST API or ZIS | High, needs developers | Fully custom, two-way | Your engineering team | Developer time plus hosting | Custom objects and multi-step, event-driven flows |
| AI-agent platform | Low to medium | Two-way, plus actions | The vendor | Usage-based, often per resolution | Resolving tickets with context from several tools |
The Zendesk Marketplace lists more than 1,800 apps, partners and integrations, by Zendesk's own count. The Zendesk REST API and the event-driven Zendesk Integration Services platform cover anything an app cannot. Direction is never uniform, so vet each one: Zendesk's Salesforce data sync, for instance, runs one way, Salesforce into Zendesk.
Choose your integration path
Before wiring anything together, match yourself to one of four paths. The right one comes down to the gap you actually feel.
Path 1: Stay native. Lean on Zendesk's native AI and a few Marketplace apps when your queue runs mostly to FAQs, password resets and basic self-service.
Path 2: Use a point connector. Pick this when a single source is missing, such as warehouse data, CRM records, form feedback or Jira issues. One connector closes it.
Path 3: Build with the REST API or ZIS. Go here for custom object logic, multi-system workflows or event-based automation that off-the-shelf apps cannot handle.
Path 4: Add an AI orchestration layer. Choose this when complex tickets need past-ticket knowledge, diagnostic questions, engineering context and a safe way to escalate. The AI layer rides on top of everything else.
High-impact use cases
These use cases shift the numbers for support teams. Each one either removes manual work or gives agents and AI more context.
- Unify customer data. Connect CRM and subscription records so support sees plan, renewal date and account value on every ticket.
- Close the feedback loop. Send tagged tickets to Jira, where product can act on feedback and feature requests.
- Bridge support and engineering. Create a linked issue from a ticket, then push status back so support can update customers without chasing anyone.
- Automate triage. Read intent, set priority and route each conversation to the right group.
- Spot trends. Roll up ticket data to find recurring issues and read them as product signals.
- Serve employees too. The same patterns drive internal help desks, so one connected stack can handle employee questions next to customer tickets.
Common approaches and where each one breaks
Every approach breaks somewhere. Spot the failure mode early to avoid an expensive rebuild.
Marketplace apps. Many stay shallow. Field coverage is partial and two-way logic is thin, so complex workflows outgrow them.
No-code tools. Fast to build, but every zap or scenario is another thing to monitor, and task-based pricing climbs with volume.
Custom API and ZIS builds. Total control, total responsibility. Without owners, docs and monitoring, webhooks fail silently and data drifts. This is a build-versus-buy call that depends on real engineering capacity.
Knowledge-first AI assistants. Most teams start here, with Zendesk's native AI, Intercom's Fin or Ada working from help center content. Simple, repeated questions go well. When the answer is outside the knowledge base, the agent stalls or escalates.
Flow-based bots. These follow keyword paths you wire by hand. Cheap and predictable, but weak on edge cases and slow to build.
AI copilots. Zendesk Copilot or Intercom Copilot draft a reply for a person to send. Trust stays high, but nothing resolves on its own.
What to look for (and when you don't need more)
Let the gap pick your integrations, not a feature list.
One customer identity across systems. Reuse one set of identifiers (account ID, external ID, email), and split organization-level fields such as plan or renewal date from user-level ones such as role or language. Botch the mapping and you get duplicate users and broken reporting.
Clear ownership. Every integration needs a named owner, one source of truth per object, and retry and error handling. Document the recovery plan before anything breaks.
Honest off-ramp #1: a clean knowledge base and mostly simple repeat questions? Zendesk's native AI agent and a couple of Marketplace apps likely cover it.
Honest off-ramp #2: moving just one kind of data? Use a point connector and stop. Census loads warehouse data into Zendesk user and organization fields; Formspree turns form submissions into tickets.
The billable unit on any AI tool. Zendesk meters AI usage in automated resolutions. Intercom's Fin bills $0.99 per outcome, counting both resolutions and procedure handoffs. Moveworks lets employees file and update Zendesk tickets inside its AI assistant. Pin down what each vendor calls a billable event before you sign.
How AI agents fit on top of Zendesk
Integrations decide what context exists in Zendesk. AI decides what happens with it. The better the inputs, the more an AI agent resolves without a person.
The ground shifted in 2026. At its Relate conference, Zendesk announced an autonomous service workforce and Model Context Protocol (MCP) support, then closed its Forethought acquisition. Zendesk is opening up to outside AI, and an agent blind to subscription status, open bugs or past resolutions will guess or escalate the moment a case turns complex.
At the simple end, AI handles triage and routing: a ticket flagged "login failure" lands in an incident queue with tighter SLAs. Next, it drafts replies and summarizes long threads, with a person on the final send. The hardest tier is resolving complex tickets end to end, and that is where an orchestration layer earns its place.
Pluno is that layer. It runs inside Zendesk, learns from your resolved tickets as well as help center articles, asks diagnostic questions to fill gaps, and reaches into the places your engineering and product teams work, such as Slack or Jira. When a person has to step in, its Escalation Copilot hands off the full ticket history with reproduction details and a suggested next step, then syncs the team's replies back into Zendesk. Pluno customers see a 65% average resolution rate and save about 48% of agent time.
"Being integrated with Zendesk directly has a completely different pool to answer questions, and then maintenance is also very easy."
Joseph D'Apuzzo, Customer Support Manager at Smartness
Best fit: technical products, troubleshooting-heavy queues and serious ticket volume, on Zendesk or any of the other help desks Pluno runs inside (Intercom, Salesforce Service Cloud, HubSpot, Jira Service Management and Freshdesk). A queue of plain FAQs is well served by a lighter knowledge-first agent. For the wider field, see Best AI Agents for Zendesk in 2026.
A safe rollout across several integrations
The five setup steps cover one app. When you are connecting several, stage them so each earns its keep before the next arrives. The timeline is a mid-market example.
Weeks 1 to 2, discovery. Inventory your tools, assign a data owner per object and pick the system of record for accounts, tickets and invoices.
Weeks 3 to 4, prioritization. Choose one to three integrations where context goes missing most often. CRM and engineering tools tend to top the list.
Weeks 5 to 8, pilot. Roll out to one group. Confirm identity mapping and track sync accuracy, adoption and resolution time.
Weeks 9 to 12, rollout. Open it to the whole team, then add AI and automation onto stable workflows.
Ongoing. Keep a catalog of integrations and owners, watch for failed syncs and recheck identity mappings each quarter.
Expected outcomes and how to measure them
Set targets before the first connection and watch the same metrics from pilot through rollout. Most sit in your Zendesk dashboards already.
- First response and resolution time. The earliest sign an integration works is tickets reaching the right owner sooner.
- Automated resolution rate. Count tickets that close without a person, and confirm each vendor's definition of a resolution.
- Agent productivity. Tickets solved per agent and time lost to searching.
- CSAT. Speed only counts if quality holds.
- Sync health. Failed events and data drift, so reporting stays honest.
Space the launches out. Ship a CRM sync and an AI agent in the same week and you cannot credit either. The ROI calculator frames the before-and-after.
What people actually say about Zendesk integrations
Zendesk admins on TrustRadius praise how quickly Marketplace apps go live, and warn about what custom and third-party builds cost to keep running.
“Plug and play apps like Round robin save us weeks of sprint delay timelines as they do not require additional coding or developer support to install and start using.”
“...we also use several integrations that allow us to bring specific context into tickets (Jira, Salesforce, Productboard)”
“A lot of these missing features require you to pay and install third party marketplace apps to fill the gaps.”
“Custom integrations can be pretty difficult to set up though.”
“Sometimes old functionality is deprecated and can break things, particularly 3rd party integrations which were very expensive to put in place”
The pattern: install apps for speed, but budget an owner for every integration, because the expensive failures come from custom builds and third-party apps that break on updates.
FAQ
What is a Zendesk integration?
A Zendesk integration is any connection that moves data or actions between Zendesk and another tool automatically. It can be a Marketplace app, a native connector, a no-code automation, a custom REST API or ZIS build, or an AI platform on top.
How do I integrate Zendesk with another tool?
Find the tool in the Zendesk Marketplace, click Install and configure its settings, or connect first-party tools in Admin Center under Apps and integrations > Integrations > Integrations. Then authorize the other tool, map the fields and test on one ticket. The five-step setup above has the detail.
Can ChatGPT or Claude integrate with Zendesk?
Partly. Since 10 August 2026, Zendesk's MCP client lets its AI action flows use tools on outside MCP servers. At Relate 2026 Zendesk also announced a Zendesk MCP Server to connect tickets and knowledge to external AI systems, and it runs an early-access program that makes ChatGPT a support channel for your customers. Until then, teams connect assistants through the API, third-party MCP apps in the Marketplace, or an AI layer such as Pluno that runs inside Zendesk.
What is Zendesk used for?
Zendesk is customer service software. Teams use it to turn email, chat, messaging, voice and social requests into tickets, route and answer them, run a help center and report on support performance. Zendesk also sells employee service for IT and HR teams.
Is Zendesk also a CRM?
Not at its core. Zendesk is a customer service platform, and its CRM, Zendesk Sell, is a separate product that Zendesk will retire on 31 August 2027, recommending Pipedrive as the replacement. Most teams connect Zendesk to a CRM such as Salesforce or HubSpot instead, using the guides in the table above.
What customer data should sync first?
Start with whatever agents open other tabs to find: plan or account tier, renewal date, recent product usage. Lock in one customer identifier before you add sources, which prevents duplicate users and keeps reporting clean.
How does Zendesk integrate with Salesforce?
The Salesforce integration syncs data one way, from Salesforce into Zendesk. Accounts become organizations, and contacts or leads become users. Our Zendesk Salesforce guide covers ticket sync and setup.
Can Zendesk integrate with Slack and Jira?
Yes. Slack handles ticket alerts and working tickets from a channel, while Jira links tickets to development issues. A common pattern links dozens of tickets to one Jira issue and updates them all when it closes.
How much do Zendesk integrations cost?
Many Marketplace apps are free or charge a monthly fee per app, and some native connectors need a specific Zendesk plan. No-code tools charge by task or run, and AI tools set their own usage models; Zendesk meters its own AI through automated resolutions. Confirm the billable unit before committing.
Do I need a developer to integrate Zendesk?
Not for Marketplace apps, native connectors or most no-code tools; you set those up in the admin UI. Custom REST API or ZIS work needs engineering, with monitoring and runbooks. AI platforms sit in the middle, connecting through secure APIs with light setup.
Next step: decide based on your tickets, not a demo
Name the gap first. Simple tickets and a current help center? A Marketplace app or two and Zendesk's native AI may be all the Zendesk integration work you need. Complex, troubleshooting-heavy cases piling onto people? That is your cue for an orchestration layer.
The best test is your own tickets. Pluno runs a free simulation on your historical tickets, so you see the quality of its answers, any gaps in its knowledge and when it would hand off, all before anything goes live.
