Our approach
We keep the data we hold small. Survey answers are stored without the respondent's name, email, IP address or device information. Everything else is protected in layers, so one mistake does not expose everything. Below is what we actually do.
Encryption and network protection
- All traffic is encrypted in transit with TLS, and HSTS is enabled.
- The site sits behind Cloudflare, which provides content delivery, DDoS protection and a web application firewall.
- Our origin server accepts connections only from Cloudflare.
Accounts and sign-in
- Passwords are hashed with a modern one-way algorithm (currently bcrypt). They are never stored or emailed in plain text. Passwords need at least 10 characters, and common passwords are rejected.
- Password reset links expire after 60 minutes and work once. Invitation links expire after 14 days. Tokens are stored only as hashes.
- Sign-in attempts are rate limited per account and per IP address.
- Session cookies are HttpOnly, Secure and SameSite. Session IDs are rotated at sign-in. Sessions end after 12 hours idle, or 14 days at most. Changing a password signs out your other sessions.
Application security
- Every form is protected against cross-site request forgery.
- A strict Content Security Policy allows scripts only from our own domain. There are no inline scripts and no third-party script tags, apart from Cloudflare's analytics beacon when it is enabled.
- Pages cannot be placed in frames, which protects against clickjacking.
- Fonts and charts are hosted by us, so pages load nothing from third-party origins.
- The database is accessed with parameterized queries only, and page output is escaped by default.
Customer separation and access
- Every record belongs to one organization, and every query is filtered by it.
- Access is role-based (admin, executive, manager, employee). Managers see only the teams they lead. Employees never see other people's answers.
Anonymity built into the data layer
- There is no stored link between a person and their survey answers. Completion is recorded separately, rounded to the hour.
- Results appear only when enough people responded. The default minimum is 4, and an administrator can set it from 3 to 10. Comments follow the same rule.
- Small teams are withheld from combined views where subtraction could reveal them.
- People without a work email can answer through a shared team link or QR code, so nothing identifies them at all.
One honest limit: comments are shown as written, so respondents are told not to include names or identifying details.
AI and data minimization
Tips come from a built-in guidance engine by default, and no data leaves the service for that. If an organization's plan includes AI-written summaries and the feature is enabled, one request per team per closed pulse goes to Anthropic's API. It contains only aggregated scores, counts, question wording, comment theme counts and at most four short comment excerpts from a group above the threshold. It never contains the organization or team name, or anyone's name or email. Flagged comments, and comments that appear to contain a name, an email address or a number, are not sent.
Logging and support access
- Administrative actions are written to an audit log.
- When our support staff open a customer workspace (a "view as" session), it is logged in that organization's audit trail and shown with a banner.
Export, deletion and backups
- Administrators on plans with exports can export results as CSV.
- On request we delete an organization's data. Our Privacy Policy gives the retention periods.
- Backups are overwritten on a rolling schedule, within about 35 days.
- The live demo uses fictional sample data only.
Health information
The service is not designed to store protected health information or patient data. Customers must not submit it. We do not sign business associate agreements at this time. If your organization needs more, talk to us.
What we do not claim
We are a small company, and we would rather be clear than impressive.
- We do not hold third-party security certifications or attestations (such as SOC 2 or ISO 27001), and we do not claim HIPAA compliance.
- We do not publish independent penetration test reports or run a bug bounty program.
- We do not promise round-the-clock monitoring, uptime percentages or service level agreements.
- We do not offer single sign-on, HRIS sync, a choice of data location, customer-managed encryption keys or audit-log export. Enterprises that need these can talk to us.
We will sign our Data Processing Addendum and answer security questionnaires on request.
Report a vulnerability
If you think you found a security problem, email [email protected] with what you saw and how to reproduce it. Please do not access other people's data, disrupt the service, or share the issue publicly before we have had a chance to fix it. We will read every report and reply as soon as we can. We do not offer rewards.
Questions
For security questions, write to [email protected] or call +1 (619) 738-1394. Our Privacy Policy and Terms of Service have the legal details.