Why Your CRM Workflow Stopped Firing (and How to Fix It Fast)
Leads aren't getting their welcome text and nobody knows why. This diagnostic playbook helps you trace a broken CRM automation to its cause in minutes, not days.
By SaaSVisionary Team · · 7 min read
It usually starts with a client comment. “Hey, we had three new leads yesterday and none of them got the text you set up.” Your stomach drops a little. The workflow was working last week. You didn’t change anything. Or did someone?
When a CRM automation stops working, the temptation is to start clicking around and changing settings until something happens. That approach occasionally works and often makes things worse. A calm, repeatable diagnostic process gets you to the answer faster and leaves a trail you can learn from.
This playbook is built for agency teams managing automations for their own pipeline and for clients. We’ll use a running example: a two-person agency in Boise that manages lead follow-up for a local HVAC company.
First, contain the damage
Before you diagnose anything, make sure nobody else falls through the cracks.
- Pull a list of affected contacts. Filter for contacts who entered the trigger condition since the problem started but didn’t receive the expected step.
- Handle them manually. Send the missed text or email by hand, or assign someone to call. Fixing the workflow can wait an hour. The leads can’t.
- Tell the client briefly. A short, honest note (“We spotted an issue with the automated follow-up, have reached out manually, and are fixing it today”) builds more trust than silence.
In Boise, that meant exporting the three leads, calling each one, and texting the HVAC owner a quick heads-up.
The five-layer diagnostic
Work through these layers in order. Each one rules out a category of problems before you move to the next.
Layer 1: Did the trigger fire at all?
Open the workflow’s execution history or activity log. Look for the affected contacts.
- If they aren’t in the log, the trigger never matched. Check the trigger conditions: form name, tag spelling, pipeline stage, or filters. A renamed form or a tag changed from “New Lead” to “new-lead” is a classic culprit.
- If they are in the log, the trigger worked. Move on.
Layer 2: Where did the contact stop?
The execution history usually shows each step a contact passed through. Find the last successful step and the first one that didn’t complete. That’s where you focus.
Layer 3: What did the failing step need?
Every step depends on something. An SMS step needs a valid mobile number, a sending number, and a working messaging connection. An email step needs a deliverable address and a connected email provider. A condition needs the field it checks to contain what you expect.
Check each dependency on one affected contact record.
Layer 4: Is an outside connection broken?
Many workflows rely on services outside the CRM: your SMS provider, email provider, calendar, payment processor, or an AI model key. Look at integration settings for:
- Expired or revoked API keys
- A password change on the connected account
- Account suspension or unpaid balance with the provider
- A disconnected calendar after someone left the team
On platforms where you bring your own Twilio, Mailgun, or AI keys, as with SaaSVisionary, a lapsed provider account is worth checking early because the workflow logic can be perfect and still fail at the send step.
Layer 5: Did something change?
If everything looks right, look at what’s different. Check the workflow’s version history if available, recent edits to forms or custom fields, new team members with edit access, and any provider-side changes such as updated carrier registration requirements.
Symptom-to-cause cheat sheet
| What you see | Likely cause | Where to look |
|---|---|---|
| No contacts entering the workflow | Trigger conditions don’t match | Trigger settings, form or tag names |
| Contacts enter but stop at a condition | Field value doesn’t match rule | Contact field, condition spelling |
| SMS steps fail, emails work | Messaging connection or registration issue | SMS provider account, number status |
| Emails show as sent but aren’t received | Deliverability or domain authentication | Email provider logs, spam folder |
| Only some contacts affected | Missing data on those records | Compare a working vs. failing contact |
| Everything stopped on a specific date | Expired key or recent edit | Integration settings, change history |
| Duplicate messages sent | Contact re-entering the workflow | Re-entry settings, overlapping triggers |
In the Boise case, the log showed contacts entering and stopping at the SMS step. The HVAC company’s messaging account had been paused over a billing issue. The workflow was fine; the connection wasn’t.
After the fix: test and verify
Don’t assume your change worked. Prove it.
- Create a test contact that matches the trigger and run it through the full workflow
- Confirm every step completes and messages actually arrive on a real phone and inbox
- Re-check the execution history for the next few live contacts
- Remove any temporary manual workarounds once you’re confident
Stop the next one before it starts
Fixing a broken workflow is good. Hearing about the next one before a client does is better.
Add monitoring
Assign an owner to every important workflow and schedule a quick weekly check of execution logs. For critical flows like new-lead response, add an alert branch that notifies the owner when a step fails. Our guide to workflow fallback paths walks through the setup.
Standardize naming
Consistent names for tags, fields, and stages prevent a surprising number of failures. Write the conventions down and share them with anyone who has edit access.
Keep a change log
A simple shared note listing “who changed what, when, and why” for each workflow turns Layer 5 from guesswork into a two-minute read.
Limit who can edit live workflows
Not everyone needs edit rights. Restricting access to a few trained people reduces accidental changes. Having CRM, messaging, and workflow automation on one platform also shrinks the number of places a connection can break.
If a lot of your client work is lead follow-up, it’s worth pairing monitoring with a solid speed-to-lead setup so gaps get caught quickly.
Frequently asked questions
Why is my CRM workflow not triggering?
The most common reason is that the trigger conditions no longer match. A form may have been renamed, a tag spelled differently, or a pipeline stage changed. Check the workflow’s activity log to see whether affected contacts appear at all. If they don’t, compare the trigger settings against the actual form, tag, or stage names on those contact records.
How do I find which step of an automation failed?
Open the workflow’s execution history and select an affected contact. Most CRMs show each step the contact reached and whether it succeeded. The first step without a success status is where you should focus. Then check what that step depends on, such as a phone number, a connected provider, or a specific field value.
What should I do for leads missed while a workflow was broken?
Contact them manually as soon as possible. Filter for contacts who met the trigger during the outage and send the missed message or call them directly. Let your client know briefly what happened and that it’s handled. Once the workflow is fixed, avoid re-running old contacts through it automatically unless you’re sure they won’t get duplicate messages.
How can I prevent CRM automation failures?
Use consistent naming for tags and fields, validate data at entry, and add alert branches to critical workflows. Assign an owner to each workflow, review execution logs weekly, and keep a simple change log. Limiting edit access to trained team members also helps, since many failures come from well-meaning changes made without knowing what depends on them.
Want automations you can monitor from one place? Start a free 14-day trial.
