Is 15Five Anonymous? What the Help Center Actually Says

✓What we say about 15Five here was last verified on 2026-08-16 against 15Five's help center and product documentation. The sources are listed at the end of this article.

01What 15Five says it does

Four statements from 15Five's documentation, which are worth holding together because each one qualifies the others:

1. Responses are linked to accounts. Engagement survey responses are associated with employee accounts within 15Five, for grouping and segmentation.
2. No individual response is surfaced. 15Five states that no individual responses are identified, and that nobody in the organization can see how a specific person responded to any survey statement.
3. Admins see participation in aggregate. During a live campaign, admins can view group-level response rates, but not individual participation status.
4. Results are withheld below a threshold. Group results aren't displayed until the confidentiality threshold is met.

Points 1 and 2 are both true at the same time, and that's precisely what "confidential" means: the link exists, and policy plus product controls keep people from following it.

02The threshold, and who can move it

This is the part worth reading closely if you're on a small team.

15Five recommends keeping the threshold at 5, and describes the smaller values as an option rather than a recommendation. Two things follow for a respondent:

• The number is a setting at your company, not a constant. Someone chose 5, 4, or 3. You're entitled to ask which.
• It's locked once the campaign starts. That's genuinely good design: the guarantee you were given at launch can't be quietly lowered after you answer.

See what changes when identity never reaches the server with the answer.

03What an employee can check before answering

None of this requires trusting a marketing page:

1. Did you have to sign in? If you reached the survey through your 15Five account, the response is attributable in the system. That's the model working as documented, not a fault.
2. Ask what the threshold is set to. "Is our confidentiality threshold 5, 4, or 3?" is a specific, answerable question, and HR either knows or can find out in feature settings.
3. Count your reporting group. If your team is 6 people and the threshold is 3, two colleagues declining to answer changes the arithmetic more than you might expect. See anonymity thresholds on small teams.
4. Watch free text hardest. Thresholds protect scores. They don't protect a writing style, a specific grievance only three people know about, or a date you mention in passing.
5. Ask who holds admin roles. The threshold is only as protective as the list of people who can change it before the next campaign.

04When confidential is enough and when it falls short

Often, yes. For a quarterly engagement pulse about workload and manager support, a documented threshold and a vendor willing to write "confidential is not the same as anonymous" on its own help pages is a reasonable arrangement, and 15Five's candour here is better than the category norm.

It stops being good enough when the topic is one where being identified carries real consequence: harassment, a safety concern, financial irregularity, whether you would leave, or how you would vote on something contested. On those questions people don't answer a confidential system honestly, and a low response rate on the sensitive items is the signal that they didn't.

For the vocabulary, see confidential vs anonymous. For a general test you can apply to any tool, see how to tell if an employee survey is anonymous and is your employee survey actually anonymous?.

05What architectural anonymity looks like instead

Disclosure: we build InviziPoll, so weigh our view accordingly.

The alternative model isn't a better setting; it's removing the link entirely so responses are encrypted in the respondent's browser and servers store ciphertext only, making individual answers unreadable by the vendor. Nothing shows until at least the poll's minimum group size has answered, which is five by default and never fewer than three.

After the first results, answers are released in random groups of three to five while at least three answers stay held in reserve until the poll closes, so no one's answer can be picked out. Breakdowns by group wait until the poll closes and need at least five people, and written answers are shown word for word, shuffled and undated, with no timestamps. Where the poll allows it, anyone who'd rather not be read can send just the topics they pick instead.

That's a narrower product than 15Five. It has no reviews, no check-ins, no OKRs. If those are why you bought 15Five, this isn't a swap. Side-by-side detail: InviziPoll vs 15Five, the wider set on the alternatives hub, and the architecture on security.

Running the survey yourself? Paste it in and see which questions could identify the people answering. Nothing you paste leaves your browser.

FAQQuestions people ask