What your field team sees is what your funder reads.
Karagem carries a programme's evidence from the enumerator's phone to the donor's desk without losing the truth on the way: offline surveys, indicators broken down by who was reached, early warning when a target is slipping, and reports in the formats funders already accept.
Field to funder, in four moves
Watch it happen on a real programme →Ask the field, properly.
Forms with skip logic, calculations and repeat groups that work without signal. Signatures, barcodes, audio and GPS captured on the spot.
Surveys, trainings and the beneficiary registry.
Trust what comes back.
Every visit is signed on the collector’s phone and sealed with its place and time, so a funder can check it without taking your word. Submissions queue for approval, duplicates get merged, and a daily scan flags out-of-range values and suspiciously fast interviews.
Verification, approvals, data quality and tasks.
See the target slipping in July.
Every indicator is forecast to programme end. Anything projected to miss shows up with the reason in plain language, before the quarter closes.
Dashboard, early warning and alerts.
Hand the funder what they asked for.
One click builds the USAID, EU, World Bank or FCDO tables from the same data, then the narrative is refined by conversation, not rewritten by hand.
Reports, scorecard and the donor dashboard.
Not screenshots. The real screens, on a demo programme.
Answer the first question and watch the form change. Yield per acre calculates itself. Nothing is submitted from this page.
Seasonal household check-in
Try it: the questions change as you answer.
What the form is doing
- Skip logic. "Which variety?" appears only after Yes; "What stopped you?" only after No. Written as XLSForm relevance, so it round-trips to Kobo and ODK.
- Constraints. An area outside 0–50 acres is refused on the phone with the message you wrote, before it can reach the dataset.
- Calculation. Yield per acre is computed from bags and area as the enumerator types; it lands in the response as its own field.
- Offline. With no signal the answer is queued on the device with any signature, photo or audio, and syncs when the network returns.
Collect it — Survey with skip logic. The form refuses what the field cannot mean — a branch that does not apply, a number outside its range — so the cleaning happens at the point of capture instead of three weeks later, from memory.
Two ways in
Where are you starting from?
Then the order matters, and it is the one most programmes get wrong: write the means of verification before the target. A target you cannot measure with the budget you have is not a target, it is a promise you will spend three years explaining.
Designing a programme from scratch is a skill of its own:Project Management in Agriculture12 weeks · Expert · USD 69.99
- 01Start from the logframe
Four levels, seven columns, one worked example. Free, no email.
- 02Define the indicators once
Unit, numerator, denominator, disaggregation — written where the data will land, not in a document nobody opens again.
- 03Set the collection up around them
Forms with skip logic that run offline, so the field can start before the network can.
- 04Reports come out of the same place
USAID, EU, World Bank and FCDO tables generated from the data, not assembled from it.
The same programme, two ways.
How it is done now
- A CSV is exported, then pasted into a workbook that holds the indicator definitions.
- Duplicates are spotted by eye, if the person doing it has had enough sleep.
- Percentages are recalculated by hand each quarter, and the formula lives in one file.
- Disaggregation sits in a tab nobody opens, so the headline number travels alone.
- A target that will miss in December is discovered in January, when it is a paragraph rather than a decision.
- The evidence for a number is a filing cabinet, a spreadsheet, and somebody’s memory.
- When that somebody leaves, the logic leaves with them.
How it is done in Karagem
- Submissions land against the indicator they feed, with no export in between.
- A daily scan flags duplicates, GPS outliers and interviews finished suspiciously fast.
- The arithmetic runs continuously, from definitions written once.
- Disaggregation is a view, not a tab — the breakdown travels with the total.
- Every indicator is forecast to programme end, so July is when you find out about December.
- A number traces to an approved submission, a person, a place and a form version.
- None of that is in anyone’s head, so nobody can take it with them.
Three people, one Monday morning.
The field officer
Achieng has 14 households to visit before the rain. Her forms work with no signal, the form refuses a farm size of 400 acres, and her afternoon task list is already on her phone.
Collect offline →The M&E officer
Brian opens the approvals queue, merges two duplicate registrations the scan caught overnight, and sees one indicator projected to miss December. He turns it into a task before 9am.
Verify and forecast →The programme director
Wanjiru's USAID quarterly is due Friday. The tables are already built from the same numbers Brian approved. She asks the report to shorten the summary and add a county comparison, and it does.
Report to funders →22
question types, from signature to repeat group
10
alert rules, delivered by email, push or WhatsApp
4
funder report templates built in
229
countries whose first-level units load with one click
2
field tools it imports from: KoboToolbox and ODK
The jargon, in the words we'd use across a table.
Monitoring and evaluation has a vocabulary that keeps new staff and small organisations out. Karagem uses the standard terms because funders do, but every one of them has a plain sentence behind it.
Logframe
A one-page table of what you will do, what should change because of it, how you'll know, and what you're assuming. Funders ask for it first; Karagem builds it on the programme page and links every row to an indicator.
Indicator
A number you commit to move: "households adopting drought-tolerant seed, from 400 to 3,000". It has a baseline, a target, a direction, and a history of recorded values.
Disaggregation
Breaking a number down by who it counts: women and men, under-35s, people with disabilities, each region. The question every funder asks second.
Baseline, midline, endline
The same survey run before, during and after a programme so the change can be measured rather than described.
Theory of change
The story of why the activities should lead to the outcome, drawn as a chain with assumptions written on the arrows. Karagem has a canvas for it.
Means of verification
Where the proof for each indicator comes from: a survey, an attendance sheet, an aggregator's ledger. Karagem records the source on every value.
Six definitions is an introduction. Eagmark teaches the whole subject:Monitoring & Evaluation for Agricultural Development6 weeks · Expert · USD 49.99
A note from the people who built it.
We built Eagmark to put agronomy, training and markets in a farmer's pocket. The organisations that use it kept asking the same thing: how do we prove to the people who fund us that any of this changed a life?
The honest answer was spreadsheets, late nights before a quarterly, and numbers nobody quite trusted. Karagem is our answer instead: one place where what the field officer records is what the director reports, with the checks in between done by the software rather than by someone at 11pm. It is built in Nairobi, priced for organisations rather than seats, and it works where the signal doesn't.
— The Eagmark team, Nairobi
Works with the tools you already use.
And the reporting standards your funders already ask for.
- KoboToolboximport forms and submissions
- ODK Centralimport forms and submissions
- Exceltemplates in, tables out
- Worddonor reports out
- WhatsAppalerts and messages to the field
- M-Pesasubscriptions in KES
- REST APIscoped keys and signed webhooks
Priced for organisations, not seats.
One subscription per organisation
Unlimited field officers. Billed through your Eagmark account in KES or USD, by M-Pesa, card or invoice.
Does it work where there is no network?
Yes. Surveys, attendance and registrations are stored on the device and sync when a connection returns. Signatures, photos and audio are queued too.
We already use KoboToolbox. Do we have to move?
No. Connect your Kobo or ODK account, import a form as a survey, and pull submissions hourly. Karagem adds the indicators, approvals, forecasting and funder reports on top.
Can our funder see progress without us sending files?
Every programme has a read-only donor dashboard on a link you can revoke. It shows reach, indicator status, budget burn and coverage, never beneficiary names.
Where is our data held, and can we take it out?
Data lives in your organisation's own space and every table exports to Excel or CSV. There is a REST API and signed webhooks for anything you want to keep in sync elsewhere.
Try it on a sample programme. Inside the real product, in under a minute.
Sign in with your Eagmark account and Karagem sets up a sample organisation for you — a Kenyan smallholder programme, as an example: sixty farmers, seven indicators, surveys with responses, tasks and alerts. Explore, then start your own in any country and any sector.