PII redaction settings

Customer feedback is full of things you want to analyze - and personal details you don’t need floating around. PII redaction strips those details out before analysis, so your team works with the substance of feedback rather than the identities inside it.

---

Where to find it

Go to Settings → PII Redaction. You first configured this during onboarding, but nothing you chose there is set in stone - this page is where you change your mind.

---

How it works

The main toggle controls automatic redaction. As the setting puts it: “We’ll remove names, emails, and other personal details before analysis.”

Redaction happens before analysis, and it applies to feedback from all sources - integrations, manual entries, imports, surveys, all of it. There’s no side door where unredacted feedback slips through into analysis.

You’ll see this in action on the mentions page: mentions that are queued for redaction show the Waiting for redaction state before moving on to analysis.

---

Per-category control

Redaction isn’t all-or-nothing. Below the main toggle you’ll find per-category switches, so you can choose exactly which kinds of PII get removed and which stay. Select all and Deselect all buttons let you flip the whole list in one click before fine-tuning.

💡 Tip: Review the category list with your privacy or legal stance in mind. The right configuration depends on what your feedback contains and what your policies require - the switches exist so you don’t have to accept a one-size-fits-all default.

---

What to expect when you change settings

A few practical notes:

• Changes apply to feedback processed going forward - redaction is a step in the pipeline that each mention passes through before analysis.
• If you disable redaction entirely, mentions skip the Waiting for redaction state and go straight to analysis.
• Redaction is independent of your feedback sources. You don’t configure it per integration; one setting governs everything coming in.

ℹ️ Note: Redaction runs before analysis by design. Sentiment, classification, and item linking all operate on the redacted text, so personal details never influence - or leak into - downstream results.

---

When to revisit this page

Most teams set redaction once and forget it. Come back when:

• You’re connecting a new kind of source whose content is more (or less) sensitive than what you’ve handled so far.
• Your privacy policy or compliance requirements change.
• Your team finds redaction is removing something they legitimately need for analysis - that’s what the per-category switches are for.

Next: Working with mentions