You do not have to leave Kobo.
Most pages like this one are an argument for switching. This one is not. KoboToolbox and ODK are very good at the thing they were built for, your field team already knows them, and a migration costs you a season. Connect your account instead: Karagem imports a form as a survey, pulls submissions hourly, and adds the layer Kobo deliberately stops short of — indicators, approvals, forecasting and the reports your funder asks for.
Karagem is an M&E platform that sits on top of data collection tools like KoboToolbox and ODK: it imports their forms, syncs submissions hourly, and turns them into indicator tracking, approval queues, forecasts and donor reports (USAID, EU, World Bank, FCDO). It has a real free tier; paid plans are priced per organisation, not per seat.Updated 2026-10-14
Where each one stops
| KoboToolbox / ODK | Karagem | |
|---|---|---|
| Offline form capture | Yes — this is what it is for | Yes, and imports your Kobo forms |
| Skip logic, calculations, repeat groups | Yes, via XLSForm | Yes, built in a visual editor |
| Submission storage and export | Yes | Yes |
| Indicators with targets and baselines | No — build it in Excel | Yes, linked to the submissions |
| Disaggregation by sex, age and location | Collected, not reported | Reported, as a first-class view |
| Approval queue before data counts | No | Yes, with a reviewer and an audit trail |
| Duplicate merging and GPS outlier flags | No | Yes, on a daily scan |
| Forecasting to programme end | No | Yes, with early warning |
| Donor report templates (USAID, EU, World Bank, FCDO) | No | Yes, one click from the same data |
| Read-only donor dashboard on a revocable link | No | Yes |
| Cost | Free for most humanitarian use | Per organisation, not per seat |
None of the “No” rows are failings. Kobo is a data collection tool and it is deliberately not a results framework — the project has said so for years. The rows are there because the work still has to happen somewhere, and right now it happens in a spreadsheet somebody maintains by hand.
The spreadsheet in the middle
Ask any organisation running Kobo how a submission becomes a number in a donor report, and the answer has a spreadsheet in it. Someone exports a CSV, pastes it into a workbook with the indicator definitions, fixes the duplicates by eye, recalculates the percentages, and sends a tab to the programme director.
That spreadsheet is where the errors live. It is also where the delay lives: it is maintained by one person, it is only as current as the last time they had an afternoon, and when they leave the organisation the logic leaves with them.
Karagem is that spreadsheet, made into software and connected directly to the submissions — so the definitions are written once, the arithmetic happens continuously, and the report is generated rather than assembled.
What connecting actually involves
Link the account
Paste a Kobo or ODK API token in Integrations. Nothing is copied out of your server; Karagem reads what the token allows and no more.
Import a form
Pick a deployed form and it becomes a Karagem survey with its questions, branches and choice lists intact. Historical submissions come with it.
Point it at an indicator
Map the question that carries the number to the indicator it feeds. From then on, every hourly pull updates the indicator, the disaggregation and the forecast.
If you actually are leaving Kobo
Some teams do migrate — usually when an organisation consolidates tools or a funder mandates one platform. The path is the same import, used once: bring the forms and the full submission history across, verify the counts against Kobo’s export, and retire the old project when the numbers agree. Your XLSForm logic survives the trip; what changes is where the data lives afterwards.
Be honest about the trade. Kobo’s free humanitarian hosting and its community are real advantages, and if pure collection is all you need, staying is the right call — the genre of page that tells you otherwise is selling something. The case for moving is everything after collection: one system holding the registry, the indicators, the approvals and the reports, instead of a collection tool plus a spreadsheet plus a document.
When you should not use Karagem
If you collect data and analyse it yourself in R or Stata, and no funder asks you for an indicator table, Kobo alone is the right answer and this would be a layer you do not need.
If you need a survey run once, for a study rather than a programme, the same is true — the value here compounds over a multi-year grant with targets to hit, and it does not accumulate in three weeks.
Where it earns its place is a programme with reporting obligations: several indicators, a target for each, disaggregation somebody checks, and a funder expecting the same tables every quarter for three years.