DocsResponses & insights
Responses, exports, and insights
Find, inspect, export, and analyze responses with schema-history context.
Work from the response grid
Open a form and choose Responses. The grid loads the newest results first and fetches older pages as you scroll, so large response sets stay usable. Columns come from the current schema plus the 200 most recent stored form versions. Removed fields within that window remain visible; a field that exists only in an older version can fall outside the working column union.
- Search runs on the server across answer values and response source metadata.
- Limit the view by date, source, respondent, or field conditions such as equals, contains, answered, and not answered.
- Use Columns to tune the working view without changing the stored response or CSV schema.
- Open a row to inspect submission context, respondent identity, files, and destination delivery state. The drawer supports previous and next navigation.
Filter first, then export
Describe the cohort
Set a date range, search term, source, respondent, or one or more field conditions. Filters are URL-backed, so the view can be revisited or shared with another authorized teammate.
Check the result count
The header count and empty state reflect the active filter. Clear or widen conditions before assuming responses are missing.
Choose CSV
The export streams every matching response, not only the rows currently loaded in the browser. It preserves the active filters and uses the current schema plus the 200 most recent stored versions for columns. A field that exists only in an older version can be absent from the CSV header even when its raw answer remains stored.
Identity columns are contextual
CSV adds respondent name, email, and id columns when that form has identified responses. Pass respondent context from your embed if downstream work needs to know who answered; see Respondents.
Inspect versions and delivery state
Every response stores the form version it answered for history and analysis. The grid and CSV build a working union from the current schema and up to 200 recent stored versions, with the latest known field definition winning. Raw older answers remain stored, but an older-only field can fall outside that union; historical labels and option wording are not guaranteed to reproduce exactly after schema changes. Sheets and Notion use the current schema when they deliver.
If a webhook, Sheets, or Notion delivery fails, the response view keeps that status next to the stored response. Fix the destination first. Retry failed selects currently loaded responses with a failure, but it requeues every destination attached to each selected response, including ones already delivered. The drawer's Re-send does the same for one response. Either path can duplicate downstream delivery.
Deletion is an operational action
Use the drawer's delete action only after checking exports and downstream copies. Deleting a Fillo response cannot retract a row or event that an external integration already accepted.
Read insights with the right confidence
Open Insights for response activity, median completion time where timing is available, source and surface breakdowns, journey completion, and question summaries. Date and field filters define the response baseline; journey health appears only in the unfiltered view and follows the selected date window. Select an answer to compare a segment with the full view.
| Measure | How to read it |
|---|---|
| Response counts and activity | Exact for the active filters and date range. |
| Question summaries and durations | Computed from the latest 10,000 matching responses when the cohort is larger; the UI marks sampled values. |
| Segment differences | Descriptive comparison, not causal analysis. Small samples are labeled directional only. |
| Journey funnel | Session progression for the selected window; use it to locate drop-off, not as a replacement for response counts. |
Move from pattern to evidence
A question card can open the matching responses. Use that filtered cohort to read the underlying answers, then export if you need offline analysis.
Separate responses from in-progress drafts
Save and resume can count open drafts and show where people stopped. Completed submissions remain in Responses. Pre-submission answers appear in a separate In progress tab only when the form owner explicitly enables that privacy-sensitive setting.
Draft recovery controls, verified identity, one-response-per-person behavior, and workspace-wide person erasure are documented in Respondents.