One conversation app. Three ways to run it.
Each type changes the consent, what is stored, and what the organisation is allowed to see. Anonymous is the flagship and ships first.
Anonymous
For sensitive, high-stakes honesty. Trust in leadership, culture, how a change actually landed, anything where the fear of being identified would otherwise flatten the answer.
The respondent follows a link, consents, talks, and leaves. No account, no download.
Paraphrased themes in aggregate only, and never who said what. No identity touches a response row, ever.
Paraphrase only, a minimum group size of five on any slice, and abuse controls on the open links.
Named private
For a closed group of lower sensitivity, where the organisation needs to know who completed the conversation but does not need attributed quotes.
The respondent follows a personal link. The report is the same aggregate as anonymous mode.
An invite layer that can show invited, started, and completed counts, so you can chase completion.
That completion reference is never joined to the answer text. The words stay unattributed in the report.
Rehearsal and release
For low-risk ideation. Pivot ideas, strategy input, pre-vetting before an offsite. For the people who would stay silent in an open forum, but may well want credit once they have seen what would be shared.
The person has a private conversation, just them and Allie.
They see a summary of what could be shared, and edit it freely.
Only then do they choose: release with my name, release without my name, or do not release. Nothing reaches the organisation until they approve it.
Side by side
Every mode uses the same session model.
A campaign carries up to three related questions in a single conversation. The agent weaves them naturally, not as a survey list. Whether you ask one question or three, each finished conversation costs one credit.
Not sure which mode fits?
Start with anonymous for anything sensitive. Talk to us if you want a second opinion.