Running Zendesk?

Zendesk triggers looping or SLAs breaching? We clear the backlog.

Zendesk gets configured by every team that touches it and ends up with 400 triggers, 70 macros, three overlapping SLA policies, and a chatbot flow nobody can read. We rationalize the automation, retire what is unused, bring SLAs back in line with the contract, and fix the integrations that drop tickets between Zendesk and Jira or Salesforce.

  • Since 2015Fixing technology
  • FreeEstimates
  • US teamReal engineers

zendesk-events.log

still broken

What it is

Zendesk at a glance

Zendesk is a customer service platform founded in Copenhagen in 2007 and headquartered in San Francisco. Its core product is Zendesk Support, a ticketing system with email, chat, voice, messaging and help center channels unified into one agent workspace. Related products include Zendesk Chat, Zendesk Talk, Zendesk Guide (knowledge base), Zendesk Sell (CRM), and the AI-centric Answer Bot and Flow Builder for self-service.

The platform is used by support, success, operations and sometimes IT teams, and it integrates with most CRMs, developer tools, and messaging platforms. Automation is handled through triggers (event-driven), automations (time-based), and macros (agent-driven canned actions), which together form the business logic of a Zendesk instance. SLAs, views and reporting through Explore wrap the lifecycle and reporting layer around the tickets.

Type
Customer service and ticketing platform
First released
2007
License
Proprietary SaaS, seat-based pricing
Maintained by
Zendesk, Inc.
Typically used for
Email, chat, phone and messaging support, help center, SLA management, customer service reporting

The honest version

Zendesk pros and cons

No technology is all upside. This is where Zendesk earns its place and where it tends to cause trouble.

Where Zendesk is strong

  • Mature omnichannel workspaceEmail, chat, phone, WhatsApp, Instagram and Facebook Messenger land in the same agent view, which is still rare at this level of polish in the segment.
  • Triggers and macros scale with the teamBusiness logic for assignment, notifications, tagging and priority can be expressed in the UI, and well-maintained macros save agents real time on repetitive tickets.
  • SLA policies are first-classFirst reply time, next reply time, requester wait time and periodic update targets can be scoped per plan, severity or customer, with breach alerts and reporting in Explore.
  • Guide is a credible knowledge baseArticles, access controls, versions and translations are built in, and search integrates back into the agent workspace to reduce repeat questions.
  • API coverage is broadTickets, users, organizations, custom fields, webhooks and audit events all have REST endpoints and most have Side Conversation and Incremental Export flavors for integrations.
  • Marketplace apps for common integrationsSalesforce, HubSpot, Jira, Slack, Shopify and many others have official apps that cover most of what a typical B2B operation needs out of the box.

Where Zendesk bites

  • Trigger sprawl becomes unmanageableWithout discipline the list of triggers runs into the hundreds, ordering becomes significant, and ticket updates start producing side effects nobody intended.
  • Reporting in Explore has a learning curveThe query builder is powerful but idiosyncratic, and real business reports often need calculated attributes that only a trained analyst can produce reliably.
  • Pricing scales with every feature addAI, advanced bots, Talk minutes and workforce add-ons each have their own meter. Mid-market plans grow in cost faster than seat count would suggest.
  • Integrations are only as good as the field mappingSalesforce, Jira and HubSpot syncs fail quietly when custom fields drift on either side. Tickets go out of sync and nobody notices until a reconciliation is run.
  • Light agent and collaboration seats have gotchasLight agents can see tickets but cannot post public replies, and the limits on what they can do surprise teams who tried to save money on seats.
  • Flow Builder is powerful but opaqueA bot flow with ten branches is fine. A flow with fifty is unmaintainable, and there is no straightforward way to diff two versions when something breaks after a change.

Sound familiar?

Common Zendesk problems we fix

The Zendesk problems that bring people to us, and what you have once each one is fixed.

Trigger loop detected on ticket 48291

Triggers fire against each other

A round-robin assignment trigger and an escalation trigger that both react to the same change on a ticket, each undoing the other and burning through rate limits.

Triggers ordered, loops broken

SLA breach: first reply > 2h

SLAs breach without anyone being paged

SLA policies written for a schedule that no longer matches the team roster, or holiday calendars out of date. The breach alert fires after the customer has already escalated.

SLAs aligned with the real schedule

Webhook to api.example.com HTTP 500

Outbound webhooks fail silently

A downstream integration that is down, a rotated API key the trigger was never updated with, or a payload schema change on the receiver. Zendesk retries and then gives up.

Webhooks delivered or alerting on failure

API rate limit 700 req/min

Integrations hit the API ceiling

A bulk sync job that hammers the Zendesk API during business hours, or an app on a lower rate-limit tier than the volume demands. Agent actions slow down.

Integrations respect the rate limits

Macro references suspended placeholder

Macros reply with raw placeholder text

A macro that references a custom ticket field which was renamed or deleted, resulting in the agent sending a reply with a literal placeholder visible to the customer.

Macros kept in step with the fields

Salesforce sync: duplicate external id

CRM integration loses ticket context

A Salesforce sync that collides on external_id because a contact was deleted and recreated, or two Zendesk organizations map to one Salesforce account. The integration bails out on a chunk of tickets.

Field mapping resilient to real data

Flow Builder: branch exceeds 10 steps

Chatbot flows become a maze

A bot flow that grew with every new question until nobody can safely edit it, and A/B comparing two versions is impossible because there is no built-in diff.

Flows split into maintainable pieces

Attachment > 50 MB rejected

Customers cannot send the files you need

Support cases that need screen recordings or large logs but which exceed the Zendesk attachment limit, so customers paste shortened versions and the diagnosis takes three extra round trips.

File handoff that actually works

What we do

Zendesk help, from one broken trigger to the whole instance

Call us for one looping trigger, or hand us the whole support operation.

Rescue
Triggers looping, SLAs breaching, or an integration dropping tickets. We stabilize the automation and then review the whole instance.
Trigger and macro cleanup
Unused triggers retired, ordering rationalized, macros reviewed against the current fields, and a documented policy for who can add new ones.
SLA and routing
SLA policies written against the real contracts, routing rules aligned with the team roster, and breach alerts that reach a human before the customer does.
Integrations
Salesforce, HubSpot, Jira, Slack and bespoke webhooks reviewed for field drift, rate-limit handling, and failure alerting.
Migrations
Moving in from Freshdesk, Intercom or Help Scout, or moving out when the fit is no longer right, with ticket history preserved.
Managed Tech
Trigger and macro audits on a cadence, SLA health reviewed each month, integration drift caught early, and a team on call when a Monday backlog threatens the plan.

How it works

From first message to fixed

  1. Look Free estimate

    Tell us what's going wrong. We look at it, then tell you plainly what's wrong and what it will cost.

  2. Fix

    We fix the cause, not the symptom. If something is down, getting it back up comes first.

  3. Harden

    Backups, monitoring, updates and access, so the same problem can't quietly come back.

  4. Hand back or stay on

    You get a plain-language write-up. Or we stay on for hosting, upkeep and new work.

Often found together

Running Zendesk alongside something else?

We work on the whole stack, so the problem can be wherever it actually is.

Questions

Zendesk questions

Something else? Call+1 737 727 7327

Why are my Zendesk triggers firing in a loop?

Two triggers each react to a change the other produces, so a single ticket update bounces between them until a rate limit or safety check steps in. The usual suspects are round-robin assignment combined with an escalation rule, or a tagging trigger combined with a categorization automation. We audit the trigger order, identify the cyclical conditions, and either split them or add guard clauses so each trigger only reacts to the change it actually cares about.

How do I stop Zendesk SLAs from breaching without warning?

Three habits catch most of it. SLA policies written against the actual support schedule and holiday calendar, not an aspirational one. Breach warning notifications set at a reasonable interval before the breach, routed somewhere a human reads. And reporting in Explore that tracks near-breaches as well as actual breaches, so the trend is visible rather than only the failure.

Our Salesforce integration keeps dropping tickets, what is wrong?

Usually a field mapping that has drifted on one side without an update on the other, a duplicate external_id from a contact that was deleted and recreated, or an API rate limit on the sync user that gets hit during the end-of-month surge. We audit the mapping against the current schema, add idempotent handlers, and separate the sync user's rate limit from the main API quota.

Can we move from Zendesk to another helpdesk without losing ticket history?

Yes. Freshdesk, Intercom, HubSpot Service Hub and Help Scout all accept ticket imports with varying fidelity. Users, organizations, custom fields, tags and attachments transfer with care. The pieces that usually need to be rebuilt are triggers, macros, SLA policies and bot flows, since those do not have clean equivalents across tools. We plan a parallel-run period before cutting over.

Is Zendesk AI and Flow Builder worth turning on?

For teams with high-volume, repetitive tickets, usually yes, but only after the knowledge base is in good shape. Answer Bot and Flow Builder deflect well when Guide articles are accurate and comprehensive, and poorly when they are thin or out of date. We tend to recommend a knowledge base audit first and automation second.

How much does Zendesk help cost?

The scoping conversation is on us. Send over the subdomain, a note on what is failing, and a screenshot of the trigger or SLA causing grief, and we send back a bounded quote.

Free estimate

Zendesk giving you trouble?

Tell us what you're running and what's going wrong. A person reads it and comes back with an estimate.

  • No charge to find out what it will take
  • If it's urgent, say so

Replies come from a person. By sending, you authorize us to email you back. No marketing.