When to graduate from KoboToolbox to an MIS
You should move beyond KoboToolbox to a management information system (MIS) when your programme depends on knowing who each record belongs to across many visits, not just what people answered on one day. The clearest signal: someone on your team spends days each month matching exports by name and village to see whether the same person improved.
This deep dive sits under our guide to KoboToolbox alternatives. If you’re new to the term, start with what an MIS for nonprofits is.
Five signals you’ve outgrown Kobo
- You link forms by hand. Registration, monthly visit and exit forms live in separate projects, and the M&E lead joins them in Excel.
- Field staff ask “who’s due this week?” and nobody can answer from the phone.
- You need statuses. Referred, enrolled, dropped out, placed, closed. Survey tools record answers; they don’t hold the state of a case.
- Managers wait for reports. Dashboards are rebuilt from exports every month, so they are always a month old.
- Funders want outcomes, not counts. “How many women increased income by 20% after 12 months?” needs one record per woman, with every visit attached.
If two or more are true, you are buying a system. A better survey app will move the pain, not remove it.
Worked example: a livelihoods programme at 3,000 participants
A composite based on programmes we’ve scoped; numbers are illustrative.
A women’s livelihoods nonprofit in Tamil Nadu enrols 3,000 participants a year across 40 villages. In Kobo it runs:
- a registration form (household details, consent, baseline income),
- a monthly visit form (training attended, savings, business status),
- an exit survey at 12 months.
Each month produces about 3,000 visit submissions, well within Kobo’s free tier. The problem isn’t volume. It is that the visit form asks for name and village, and the M&E officer spends four to five days a month matching visits to registrations. About 8% of visits can’t be matched because names are spelled differently [illustrative]. The donor report on income change is therefore built on the 92% that match, and nobody is sure the missing 8% are random.
What the MIS changes:
| Today in Kobo | In a programme MIS |
|---|---|
| Name + village typed on every form | Participant selected from a list; unique ID carried automatically |
| Three separate projects | One participant record with registration, 12 visits and exit attached |
| Monthly Excel matching | No matching; the record is already joined |
| Field officer can't see history offline | History and due follow-ups on the phone, offline |
| Report built monthly from exports | Dashboard by village, cohort and field officer, updated as data syncs |
The design work starts from the Theory of Change, not the form. Our Theory of Change to data model guide shows how to turn outcomes into records, fields and indicators. For a programme that serves individuals, the core record design is covered in our guide to a beneficiary management system.
Three migration paths
1. Keep Kobo, add a system behind it. Kobo stays the field tool; submissions sync to a database through Kobo’s API or a scheduled export, and the system handles IDs, history and dashboards. Cheapest and least disruptive. Works when follow-ups don’t need history on the phone.
2. Partial move. Registration and follow-ups move to the MIS (Zoho Creator, CommCare or Avni) so field staff can select existing participants offline. One-off surveys such as baselines and perception studies stay in Kobo. This is the most common pattern in our work: in a disaster-response engagement we scoped, the field assessment step stayed in Kobo while programme, volunteer and donor work moved to Zoho Creator.
3. Full rebuild. Everything moves to one platform. Right when you need approvals, referrals and case statuses everywhere, or a single access-control and consent model across all data.
Before you graduate: a one-page readiness check
- Do we have one agreed participant or household ID format?
- Have we written five to ten indicators the system must report, taken from our Theory of Change?
- Do we know who will own the system after go-live, by name?
- Have we checked consent and data location for personal data? Use our DPDP compliance checklist if you work in India.
- Have we budgeted training and three months of support, not just the build?
If you can’t answer the first two, spend a month on them before you buy anything. If you are still deciding which survey tool to keep for one-off studies, our KoboToolbox vs SurveyCTO comparison will help.
Not sure whether you need a new tool or a new system? Take EdZola’s free 10-minute technology assessment. It tells you whether to tune Kobo, switch tools, or move to a programme system, and what to fix first.