Does it work in villages without mobile signal?

Offline collection is coming at launch: interviews will save on the device and sync when the phone is back in coverage, using the one-page form, because the follow-up question needs a connection to be written. Today the survey needs a connection, and runs as a short chat with follow-ups or as a one-page form.

Can respondents who can't read take part?

Yes. The enumerator reads each question aloud and records the answer, or the respondent speaks it into the phone in their own language on phones that support voice input. Nobody needs an account or an app of their own. You read every answer translated into English, with the original words kept beside it.

Can we collect health information, and where is it stored?

Responses are stored in the EU (Frankfurt) or the US, set from where you sign up. Health information is special-category data under GDPR, and our Terms allow collecting it when you have a lawful basis and the respondent's informed consent, which you can record on a consent screen before question one. You set how long responses are kept, can delete one household's answers on request, and the processors are listed on our subprocessors page and in the data processing agreement.

Is it accurate to let AI sort what respondents said?

Every theme shows roughly how many households said it and opens to example answers, each next to the respondent's own words, so your team can check any summary. Moving answers, renaming themes and coding into your own categories are coming at launch. The follow-up is one neutral question about the respondent's own answer, asked only when an answer is too vague to use, and you can switch it off for any question.

How much does one survey round cost?

Joining the waitlist is free. A baseline or endline round fits the Event Pass: $79 once for 60 days and 3,000 responses. Starter is $19 a month for 500 responses with themes and CSV export, Pro is $49 a month for 3,000 responses, and Business is $149 a month with five staff seats. Enumerators and respondents need no seat or account. Every plan starts with a 7-day trial and is paid by card.

How is this different from KoboToolbox or ODK?

KoboToolbox has a free Community plan and paid tiers, ODK is open source, and both are built for long offline questionnaires with closed questions. Kobo now analyses audio answers with AI. FormzoAI asks one follow-up when an answer is a single word, translates every answer into English, and groups the answers into counted themes with quotes. You can keep Kobo for the closed questions and run the open ones in FormzoAI.

For NGO and M&E field teams

486 household interviews, every open answer read and counted.

When a mother says “distance”, FormzoAI asks what happens when a child gets sick at night. Every answer comes back in English, in counted themes with quotes for the donor report.

Household health survey

Kilifi clinic access endline · made from “Endline household survey on clinic access in Kilifi, 40 villages, Swahili and English, enumerators on shared phones”

Their chat

What makes it hard to get care for your children?

Distance.

Follow-upWhat happens when a child gets sick at night?

The clinic is 14 km away. We wait for a boda boda until morning, and the ride costs 600 shillings.

What you get

Theme

No transport after dark

211 of 486 people said something like this

“The clinic is 14 km away. We wait for a boda boda until morning, and the ride costs 600 shillings.”

Example answers, written for this page.

Week three of data collection.

“Distance.” “Money.” “Nothing.”
By item 63 of 80 the enumerator types one word and moves on. The endline report has to say why families skip the clinic, and the open column says “distance” 300 times.
Six languages, one paper translation
Enumerators translate the question on the doorstep, each a little differently, then write the Giriama answer down in English shorthand. Nobody at the office can check what was really said.
1,400 open answers in Excel
The export lands the week before the donor report. Someone on the M&E team codes every row by hand, and the qualitative section ends up as three quotes and a guess.

From one sentence to the endline report.

01
Describe the survey in one sentence
“Endline household survey on clinic access in Kilifi, 40 villages, Swahili and English.” You get the questions with a consent screen first, and your team can reword any item before training day.
02
One phone per enumerator
Open the link with the enumerator's ID in it, and walk. Shared-device mode clears the form after each household, and respondents answer in their language, typed or spoken. Saving offline and syncing when the phone finds signal is coming at launch.
03
Write the report from the themes
Every open answer translated into English and grouped into themes, each with how many households said it and the respondent's own words beside it. Export the table as CSV, with village and enumerator on every row. A themes report with a method note is coming at launch.

What 486 households said about reaching the clinic.

Answers are grouped into themes, each with how many households said it and its best quote, translated, with the original Swahili beside it. Coding into your own codebook, like transport, cost or quality of care, is coming at launch.

Kilifi clinic access endline · households

0 in

Built for the doorstep and the donor report.

Every language in the district: Respondents answer in Swahili, Giriama or whatever they speak at home. You read every answer in English, with their original words kept beside it.
Answer by speaking: Tap the microphone and talk, on phones that support voice input. Useful for respondents who can't read, and faster than the enumerator typing.
One phone, many households: Shared-device mode clears the form after each interview, and the enumerator's ID on the link tags every answer, so you can see who collected what.
Collect without signal: Interviews save on the phone in the village and sync when the enumerator is back in coverage. Coming at launch.
A link per village: Add ?village=mtwapa to the link and every answer carries it, so you can compare villages in the export.
Counts for the logframe: Each theme shows roughly how many households said it, like no transport after dark, with example answers a click away. Coding into your own categories is coming at launch.

Fair questions.

Does it work in villages without mobile signal?: Offline collection is coming at launch: interviews will save on the device and sync when the phone is back in coverage, using the one-page form, because the follow-up question needs a connection to be written. Today the survey needs a connection, and runs as a short chat with follow-ups or as a one-page form.
Can respondents who can't read take part?: Yes. The enumerator reads each question aloud and records the answer, or the respondent speaks it into the phone in their own language on phones that support voice input. Nobody needs an account or an app of their own. You read every answer translated into English, with the original words kept beside it.
Can we collect health information, and where is it stored?: Responses are stored in the EU (Frankfurt) or the US, set from where you sign up. Health information is special-category data under GDPR, and our Terms allow collecting it when you have a lawful basis and the respondent's informed consent, which you can record on a consent screen before question one. You set how long responses are kept, can delete one household's answers on request, and the processors are listed on our subprocessors page and in the data processing agreement.
Is it accurate to let AI sort what respondents said?: Every theme shows roughly how many households said it and opens to example answers, each next to the respondent's own words, so your team can check any summary. Moving answers, renaming themes and coding into your own categories are coming at launch. The follow-up is one neutral question about the respondent's own answer, asked only when an answer is too vague to use, and you can switch it off for any question.
How much does one survey round cost?: Joining the waitlist is free. A baseline or endline round fits the Event Pass: $79 once for 60 days and 3,000 responses. Starter is $19 a month for 500 responses with themes and CSV export, Pro is $49 a month for 3,000 responses, and Business is $149 a month with five staff seats. Enumerators and respondents need no seat or account. Every plan starts with a 7-day trial and is paid by card.
How is this different from KoboToolbox or ODK?: KoboToolbox has a free Community plan and paid tiers, ODK is open source, and both are built for long offline questionnaires with closed questions. Kobo now analyses audio answers with AI. FormzoAI asks one follow-up when an answer is a single word, translates every answer into English, and groups the answers into counted themes with quotes. You can keep Kobo for the closed questions and run the open ones in FormzoAI.

Get in before your next survey round goes to the field.

Get early access

Free to join. Spots open in batches.

486 household interviews, every open answer read and counted.

When a mother says “distance”, FormzoAI asks what happens when a child gets sick at night. Every answer comes back in English, in counted themes with quotes for the donor report.

Household health survey

Kilifi clinic access endline · made from “Endline household survey on clinic access in Kilifi, 40 villages, Swahili and English, enumerators on shared phones”

Their chat

What makes it hard to get care for your children?

Distance.

Follow-upWhat happens when a child gets sick at night?

The clinic is 14 km away. We wait for a boda boda until morning, and the ride costs 600 shillings.

What you get

Theme

No transport after dark

211 of 486 people said something like this

“The clinic is 14 km away. We wait for a boda boda until morning, and the ride costs 600 shillings.”

Example answers, written for this page.

Week three of data collection.

“Distance.” “Money.” “Nothing.”

By item 63 of 80 the enumerator types one word and moves on. The endline report has to say why families skip the clinic, and the open column says “distance” 300 times.

Six languages, one paper translation

Enumerators translate the question on the doorstep, each a little differently, then write the Giriama answer down in English shorthand. Nobody at the office can check what was really said.

1,400 open answers in Excel

The export lands the week before the donor report. Someone on the M&E team codes every row by hand, and the qualitative section ends up as three quotes and a guess.

From one sentence to the endline report.

Describe the survey in one sentence

“Endline household survey on clinic access in Kilifi, 40 villages, Swahili and English.” You get the questions with a consent screen first, and your team can reword any item before training day.

One phone per enumerator

Open the link with the enumerator's ID in it, and walk. Shared-device mode clears the form after each household, and respondents answer in their language, typed or spoken. Saving offline and syncing when the phone finds signal is coming at launch.

Write the report from the themes

Every open answer translated into English and grouped into themes, each with how many households said it and the respondent's own words beside it. Export the table as CSV, with village and enumerator on every row. A themes report with a method note is coming at launch.

Built for the doorstep and the donor report.

Every language in the district

Respondents answer in Swahili, Giriama or whatever they speak at home. You read every answer in English, with their original words kept beside it.

Answer by speaking

Tap the microphone and talk, on phones that support voice input. Useful for respondents who can't read, and faster than the enumerator typing.

One phone, many households

Shared-device mode clears the form after each interview, and the enumerator's ID on the link tags every answer, so you can see who collected what.

Collect without signal

Interviews save on the phone in the village and sync when the enumerator is back in coverage. Coming at launch.

A link per village

Add ?village=mtwapa to the link and every answer carries it, so you can compare villages in the export.

Counts for the logframe

Each theme shows roughly how many households said it, like no transport after dark, with example answers a click away. Coding into your own categories is coming at launch.

Fair questions.

Does it work in villages without mobile signal?: Offline collection is coming at launch: interviews will save on the device and sync when the phone is back in coverage, using the one-page form, because the follow-up question needs a connection to be written. Today the survey needs a connection, and runs as a short chat with follow-ups or as a one-page form.
Can respondents who can't read take part?: Yes. The enumerator reads each question aloud and records the answer, or the respondent speaks it into the phone in their own language on phones that support voice input. Nobody needs an account or an app of their own. You read every answer translated into English, with the original words kept beside it.
Can we collect health information, and where is it stored?: Responses are stored in the EU (Frankfurt) or the US, set from where you sign up. Health information is special-category data under GDPR, and our Terms allow collecting it when you have a lawful basis and the respondent's informed consent, which you can record on a consent screen before question one. You set how long responses are kept, can delete one household's answers on request, and the processors are listed on our subprocessors page and in the data processing agreement.
Is it accurate to let AI sort what respondents said?: Every theme shows roughly how many households said it and opens to example answers, each next to the respondent's own words, so your team can check any summary. Moving answers, renaming themes and coding into your own categories are coming at launch. The follow-up is one neutral question about the respondent's own answer, asked only when an answer is too vague to use, and you can switch it off for any question.
How much does one survey round cost?: Joining the waitlist is free. A baseline or endline round fits the Event Pass: $79 once for 60 days and 3,000 responses. Starter is $19 a month for 500 responses with themes and CSV export, Pro is $49 a month for 3,000 responses, and Business is $149 a month with five staff seats. Enumerators and respondents need no seat or account. Every plan starts with a 7-day trial and is paid by card.
How is this different from KoboToolbox or ODK?: KoboToolbox has a free Community plan and paid tiers, ODK is open source, and both are built for long offline questionnaires with closed questions. Kobo now analyses audio answers with AI. FormzoAI asks one follow-up when an answer is a single word, translates every answer into English, and groups the answers into counted themes with quotes. You can keep Kobo for the closed questions and run the open ones in FormzoAI.