Troubleshooting

Working through a problem? Each entry below gives the symptom, the likely cause, and the fix. If your issue isn’t here — or the fix doesn’t work — contact support.

Voice login shows an error (503)

Symptom: When enrolling or verifying your voice, you see an error (often a “503 Service Unavailable” or a message that voice isn’t available).

Likely cause: Voice login isn’t currently configured or enabled on the account you’re using — this is the system telling you the feature isn’t set up, not that something broke with your voice.

Fix: Check your account settings for the voice login option. If voice login isn’t offered there, it isn’t enabled for your account yet — use your normal password login (with MFA) and contact support if you were expecting voice to be available.

Survey responses aren’t syncing

Symptom: You’ve collected survey responses, but they’re not showing up in your results.

Likely cause: The survey delivery integration is disconnected — most commonly an expired or revoked connection token.

Fix: Go to the integration settings, reconnect the survey platform, and re-run the sync. After reconnecting, new responses should flow through normally. If responses are still missing after a reconnect, contact support.

Landing page deploy fails

Symptom: You try to deploy your landing page and the deployment fails or never goes live.

Likely cause: A problem with the deployment target — the page content is usually fine; the connection to the hosting destination is what’s failing.

Fix: Check that your hosting credentials and destination are still valid, then retry the deploy. If it fails again, contact support with the page name and the error text.

Research results look “generic”

Symptom: A research run returns results that read like obvious, generic statements with no specifics about your idea.

Likely cause: A research or AI data source isn’t configured for that feature, so the run produced placeholder or simulated content instead of live data.

Fix: Check the results — simulated content is always clearly labeled as simulated, never presented as real data. Configure the relevant data source (or upgrade to a plan that includes it), then re-run the research. If you believe a real source should be active but isn’t, contact support.

A research run fails to complete

Symptom: A research run stays stuck, errors out partway, or never produces a report.

Likely cause: One of the sources the run depends on was temporarily unavailable, or the run was interrupted.

Fix: Restart the research run from the idea page — most runs complete on the retry. If the same run keeps failing at the same point, contact support and mention which research phase is failing.

AI responses are slow

Symptom: AI-generated content or research takes noticeably longer than usual.

Likely cause: The AI model is busy or the request is large — long prompts and multi-part research naturally take more time than short ones.

Fix: Usually just a matter of waiting. If responses are consistently slower than before, try again a few minutes later. For ongoing slowness, contact support.

I see “plan limit” messages after downgrading

Symptom: After your trial ended or you downgraded, you see messages that a feature is locked or a limit is reached.

Likely cause: That’s expected — the feature or allowance belongs to a higher plan, and your account is now on the Free (or lower) plan.

Fix: Review the Pricing page to see which features your current plan includes. You can upgrade any time to unlock them again; your data is never deleted by a downgrade.

Still stuck?

If none of these match your situation, contact support — open the support widget in the product and choose Still need help?, or email support directly. Include what you were doing, the exact error message, and when it started.