Troubleshooting
Solutions to the problems teams run into most often, with the exact checks to run. If none of these help, contact support with your workspace ID and the steps you've tried — it saves a round of questions.
On this page
No data after installing#
If Settings → Tracking shows no events a minute after you've visited your site, work through these checks in order:
- Disable ad blockers and privacy extensions in the browser you're testing with, or test in a clean private window. Some block any script with “tracking” in its path.
- Confirm the snippet is on the page. View the page source and search for
cdn.aurora.io. If it's missing, the template you edited isn't the one used for that page, or a cache needs clearing. - Check the workspace key. The
data-workspacevalue must match the key shown in your dashboard exactly, including thepk_live_prefix. - Look for console errors. A Content Security Policy error means you need to allow
cdn.aurora.ioandcollect.aurora.io— see Content Security Policy. - Check consent mode. If the snippet has
data-consent="pending", nothing is sent untilaurora('consent', 'granted')is called. - Check filters. Your own IP or email domain may be excluded under Settings → Tracking → Filters.
Page views counted twice#
This almost always happens in single-page apps where both data-spa="true" and a manual aurora('page') call in the router are active. Use one or the other. It can also happen when the snippet is included in both a GTM tag and the page template — remove one.
Scores look too high or too low#
- Everyone qualifies: a rule probably matches too broadly — for example Starts with
/instead of/pricing, or a rule set to score every time. Open a qualified visitor and look at the contributing rules. - Nobody qualifies: check that your URL patterns match your real paths, including locale prefixes such as
/en/pricing, and that custom event names match exactly. The rule preview shows zero matches when a pattern is wrong. - Scores dropped overnight: weekly score decay reduces inactive visitors by 10%. This is intended; adjust it under Scoring → Settings if your sales cycle is longer.
For a structured approach to tuning, see Choosing a threshold.
A qualified lead didn't trigger an alert#
- Open the visitor's timeline. It shows which routing rule matched and which destinations were notified, or why none were.
- If the lead qualified during a re-score, no alert is sent by design. It appears in the shared inbox instead.
- Check quiet hours and the assignee's daily alert limit; held alerts are delivered later as a summary.
- If the lead was already routed earlier, later activity is posted as a thread reply rather than a new alert.
Slack or Teams alerts stopped#
If alerts stop suddenly, the app was usually removed from the channel, the channel was archived, or the person who connected the integration left your Slack workspace. Go to Integrations → Slack: a warning explains the problem. Reconnect with an account that will stay in the workspace, and for private channels run /invite @Aurora again.
CRM sync errors#
Failed records are listed on the integration page with the error returned by your CRM. The most common causes are:
| Error | Cause and fix |
|---|---|
REQUIRED_FIELD_MISSING | A field required on your Lead or Contact layout has no value. Map a default value for it, or relax the requirement for the integration user. |
FIELD_CUSTOM_VALIDATION_EXCEPTION | A validation rule in Salesforce rejected the update. Exempt the integration user or adjust the rule. |
INVALID_OR_NULL_FOR_RESTRICTED_PICKLIST | Aurora wrote a status your picklist doesn't allow. Add the value or remap statuses. |
UNABLE_TO_LOCK_ROW | Another process was editing the record. Aurora retries automatically within an hour. |
HubSpot 429 | Your HubSpot account's rate limit was reached. Updates are queued and resume automatically. |
SSO sign-in fails#
- “Invalid SAML response”: the ACS URL or Entity ID in your identity provider doesn't match Aurora exactly, or you're using the US URL for an EU workspace.
- “User not assigned”: assign the Aurora application to the user or their group in your identity provider.
- “Email attribute missing”: map the
emailattribute and set the Name ID format toEmailAddress. - Clock errors: assertions are rejected if the identity provider's clock is more than three minutes off. Check its time synchronization.
Locked out after enforcing SSO? Owners can still sign in with their recovery login at app.aurora.io/login?recovery=1.
Webhook deliveries failing#
Open Settings → Webhooks → Delivery log to see the exact response from your endpoint for each attempt.
- Timeouts: your endpoint took longer than 10 seconds. Return
200immediately and process the event in the background. - Signature mismatch: verify against the raw request body, not re-serialized JSON, and use the secret for that specific endpoint.
- TLS errors: the endpoint's certificate is expired, self-signed or missing intermediate certificates.
- Endpoint disabled: endpoints that fail for three consecutive days are disabled. Fix the cause, re-enable the endpoint and redeliver failed events from the log.
Contacting support#
Include your workspace ID (Settings → General), the visitor ID or request ID involved, what you expected and what happened instead, and screenshots of any errors. Send it through the contact page or to support@aurora.io. Check the status page first if several things stopped working at once.