A research instrument worthy of the work.
Plan your instrument, collect mixed-methods data offline, then code the open responses with Margin — a research AI grounded in your own dataset. Built for the researchers, NGO M&E teams, and consultancies who actually have to make sense of what they gather.
"The coordination between us and the Births & Deaths Registry has improved, but regional offices still operate on three different data standards. Until that's resolved, integration is more on paper than in practice."
Three audiences. One tool.
We didn't build for "everyone." We built for the people who actually have to collect qualitative + quantitative data and then make sense of it.
NGOs & M&E teams
Baseline-midline-endline surveys, programme evaluations, dialogues. Audio interviews, GPS, photos — collected offline by a roster of enumerators, analysed centrally.
Academic researchers
PhD fieldwork, faculty studies, mixed-methods theses. Versioned instruments, ethics-aware consent, exports for SPSS / Stata / NVivo. Plus Margin to help theme open-text answers.
Consultancies
Short engagements, multiple clients, polished outputs. Branded instruments, public response links, narrative reports drafted from your data — without Excel-stitching weeks.
The whole loop — designed, not duct-taped.
KoboToolbox for quant. Otter for transcripts. NVivo for coding. Excel for cross-tabs. Dropbox holding it together. Fieldnote replaces that stack with one product.
Drag-build instruments with text, Likert, GPS, audio — one form, mixed methods.
Versioned automatically. Old responses stay tied to the version they were collected on. Multi-language, skip logic, calculated fields, repeating groups.
Offline-first PWA — install once, work for days without a signal.
Service worker caches the app shell, IndexedDB stores responses locally, sync resumes when the connection returns. Tested across 3G, EDGE, and zero-bar deployments.
Append-only, idempotent, conflict-aware.
Every response carries a client UUID. Submit it twice — Fieldnote accepts it once. No silent overwrites, no lost interviews, full audit log.
An AI that reads your responses, cites the ones it quotes, and never invents data.
Margin is your in-product research analyst. Summarise sets of responses, extract themes from open text, draft narrative paragraphs — all grounded in the actual data, with citations.
Map-based collection — drop pins, capture tracks, click pre-loaded features.
GPS points today; polygons, lines, and tile-served basemaps next. The map is the input control where location is the answer.
Send public response links by email — no app install required.
For expert interviews and online surveys, paste a list of recipients and Fieldnote generates per-respondent links. Track sent, opened, submitted.
Three steps, one continuous flow.
From the moment you draft an instrument to the moment you read your final report, the data never leaves Fieldnote. Nothing gets lost in CSV land.
Design
Draft an instrument in the visual designer. Add field types, skip logic, validation. Publish it — Fieldnote freezes that version forever; future edits fork a new draft.
Collect
Enumerators install the PWA, sync once, and go offline. Or send anonymous public links by email. Either path lands in the same response table.
Analyse
Dashboards, cross-tabs, map views, AI summaries via Margin. Export to CSV, GeoJSON, SPSS, Stata — or pipe directly into your own database.
Your research analyst — calm, exact, and never out of context.
Margin lives in the margin of your fieldnote. Ask her to summarise a hundred interviews, find the themes across regions, draft your introduction paragraph, or check whether a specific quote came from your dataset. Every claim cites the responses it came from.
See Margin in action ›Stop stitching tools together. Start collecting like you mean it.
Solo plan, team plan, or organisation plan. No free tier — but no surprises either. Get a workspace seeded with sample data on signup.