The six weeks between ethics approval and your first survey wave are a strange period. The instrument is ready. The participant list is built. The field logistics are locked. And someone still has to figure out how to actually reach 400 farmers in western Kenya and northern Tanzania on a regular basis for the next 18 months, ask them structured questions, pay them something for their time, and bring back data clean enough to analyze.
Mobile phone surveys for field research are the practice of using SMS, WhatsApp, or voice to administer structured surveys to a defined participant panel across one or more countries, typically with a digital incentive, on a recurring schedule tied to a study's data collection rounds. They are not a replacement for in-person baseline work. They are what you deploy between field visits to maintain panel contact, collect interval data, and keep attrition low enough that your endline is still statistically meaningful.
This post covers the practical setup: channel choice, participant recruitment and consent, data storage and export, incentive design and payment, ethics mechanics, and how to be ready to send your first wave in the same window you are finalizing your survey instrument. It is written for the researcher who will operate the system, not the PI whose grant pays for it.
Which channel fits your study
The channel question is real and the answer depends on your participant population. It is worth making the decision deliberately rather than defaulting to whatever the platform demo showed.
SMS is the right baseline for most field research applications. It reaches any phone on any network with no data cost to the participant, no app install, and no smartphone requirement. Delivery rates are verifiable, response data is clean text, and the workflow runs identically whether the participant is in a market town or a rural area with intermittent data coverage. For longitudinal panels in sub-Saharan Africa and South Asia, SMS is the channel that keeps attrition from becoming a network-access problem on top of everything else.
WhatsApp works well for enumerator coordination, for respondents in urban areas where smartphone penetration is high, or as an alternative channel for participants who do not respond to SMS. It should not be your primary instrument delivery channel in rural field sites where feature phone use is still common. A well-built survey deployment checks whether a given number has WhatsApp and uses that where it exists, then falls back to SMS for the rest, without your team making that determination manually for each contact.
USSD is worth knowing exists even if it is not your primary tool. It runs on feature phones, costs the participant nothing, and works in places where SMS delivery can be inconsistent. The constraint is that it is session-based: the participant has to respond in real time during the USSD session, which changes your survey design. What USSD is and when it applies deserves a separate read if your field site has significant feature phone density or unreliable SMS delivery.
For a deeper comparison of channel options by respondent profile, the channel selection guide is the cleaner reference. Make the decision once, document it in your methods, and configure accordingly.
Getting local sending presence without a local entity
The practical problem for multi-country studies is the same one every new field team runs into: you need a local phone number or sender ID in Kenya, a different one in Tanzania, and ideally both before you arrive in-country for your baseline visit.
The fastest path is the Android Gateway. An Android phone with a local SIM and the Telerivet gateway app installed becomes a two-way SMS route connected to your account, same day, with no carrier API contracts and no in-country office. You need a local partner, enumerator, or colleague who can keep the device plugged in and connected to data. Once it is running it operates as a persistent gateway: your survey messages go out through that local number, replies come back into your Telerivet account in real time, and your participant receives a message from a number they recognize as local. For the scale of most academic field studies (a few hundred to a few thousand participants), one device per country is sufficient. The Android Gateway field guide covers setup in detail.
If your study has an in-country partner organization with existing aggregator relationships, those routes connect through Telerivet's API and your team handles only the study logic, not the telecom plumbing.
One misconception to clear up early: the platform does not supply phone numbers for participant recruitment. You need a pre-existing contact list. The platform sends to numbers you already hold.
Building the panel, managing consent
Consent for mobile outreach in field research runs through two separate bodies: your home institution's IRB and the host country's equivalent. The Hebrew University of Jerusalem researcher who raised this with Telerivet directly named Kenya's ERC and NACOSTI in the same breath. Both matters. Getting institutional consent at home does not substitute for compliance with the data protection framework governing your field site.
The platform mechanics that support ethics compliance are: an opt-in confirmation message sent to every participant before the study begins, explicit opt-out language included in every survey invitation, and the ability to remove a contact from all future outreach with one action. These are configurable defaults, not afterthoughts. The ethics board's specific consent language goes into the first outbound message template. What happens after that is a study design question, not a platform question.
Participant data lives in Telerivet's contact management system. You can see every contact, every response, and every message exchange in real time from anywhere. Export to CSV or Excel is available on the Pro plan at any point, and the data structure is clean enough to pull directly into R or Stata. If your data management plan requires a specific schema or integration with a survey platform like SurveyCTO or KoBoToolbox, the REST API and webhook connections handle that.
Survey design for mobile, and the incentive gaming problem
Three to five questions per wave is the practitioner consensus. The airtime incentive meaningfully increases tolerance for longer instruments, but the design principle still holds: ask only what you cannot get another way, keep language simple enough for a first-pass SMS read, and test the full flow with a small group before sending to the full panel.
The incentive gaming problem is worth addressing directly because it comes up in every study using airtime rewards and has a methods answer rather than a platform answer. If participants learn that answering yes to everything earns airtime quickly, response quality drops in a way that is visible only in analysis, after the fact. The countermeasures are: avoid binary yes/no formats wherever branching or multiple choice applies, include at least one free-form question per wave where effort level is legible, gate incentive delivery to pre-approved contacts in your panel rather than any incoming number, and vary the incentive amount based on completion rather than speed. Telerivet supports all of these configurations natively. The point is to design the incentive so that the fastest path to the reward is also the path of a genuine response.
For more on mobile incentive design in development contexts, the mobile rewards implementation guide covers the broader pattern, and the airtime as incentive piece covers the public health evidence on effectiveness.
Paying the incentive
Airtime delivery is available natively across hundreds of mobile networks in Africa and Asia. A participant completes the survey wave, the workflow confirms they have reached the final question, and the airtime transfer triggers automatically. No manual reconciliation, no separate payment system, no in-country cash logistics. For panel studies running quarterly waves over 18 months, that automation is the difference between an incentive program that runs itself and one that becomes a coordination task after every wave.
Airtime requires the Pro plan. The platform fee is itemized separately from the message transport costs and the airtime pool itself. For a 500-person panel with a small per-wave incentive, the airtime cost is typically the largest line, but it is still a fraction of what a single in-person data collection round costs in travel and field staff time.
Digital vouchers and other reward types are also available where the network supports them. For most field sites in East Africa and the Philippines, airtime is the right default: it has no expiry, transfers instantly, and is valued uniformly across income levels and locations.
Panel attrition and response tracking
Attrition in longitudinal mobile panels follows predictable patterns: number changes, network switching, and non-response in the second wave by contacts who participated once. None of these are avoidable entirely. They are manageable if you are watching the data.
Telerivet's data tables track individual contact response status per wave, so you can see attrition in real time rather than at analysis. Contacts who do not respond to a wave can receive a single follow-up message before the wave closes, without sending the full instrument again. Contacts who miss two consecutive waves can be flagged for an enumerator call or removed from the panel, depending on your protocol. This is the tracking layer that lets you maintain a panel long enough for a meaningful endline.
For situations where SMS does not reach a contact and a zero-cost alternative exists, missed-call polling is worth knowing: a participant misses-calls a number to register a response, incurring no data or airtime cost, and the inbound call trigger routes into the same workflow. It is a niche application but a genuine one for populations where even small message costs are a barrier.
What this looks like to an operator
The practical setup for a two-country longitudinal study: one Telerivet project with two Android Gateway routes (one local SIM per country), a participant contact list imported from the field, consent message sent as the first outbound, a poll service built for each wave, airtime reward triggered on survey completion, response data exportable to CSV after each wave. Accessible in real time from a laptop. No engineering support required. The full SMS polling how-to walks through the mechanics.
If you have used Telerivet before with a different organization, the mobile surveys guide written for NGO teams is the closest analogue to the academic field research workflow. The setup is the same. The compliance context differs in the details.
Frequently asked questions
Do I need technical skills to set up a mobile survey in Telerivet? No. Survey waves, opt-in flows, and incentive delivery are all configurable through the platform interface. If you can describe the survey in plain language, you can build it. Telerivet also supports MCP integration, which means you can instruct an AI assistant to build and deploy a survey by describing it conversationally, with your confirmation before anything sends.
Can Telerivet provide a list of participants for my study? No. The platform sends to contact lists you supply. Telerivet does not maintain a participant panel for external recruitment.
How do I handle participants in two countries who use different mobile money networks? Airtime delivery works independently per network. Telerivet routes each transfer to the correct network automatically based on the contact's number. You do not manage this per participant.
Does the platform handle ethics and IRB compliance? The platform supports the mechanics: opt-in confirmation, opt-out handling, consent language in outbound templates, and data export for your records. Ethics approval with your institution and with host-country bodies is the researcher's responsibility. Telerivet does not review study protocols or supply compliance documentation.
What does a typical study cost per month on Telerivet? Most field research studies fit the Pro plan at $149 per month. Message transport costs (what carriers charge per SMS) and airtime incentives are separate and priced at cost, without markup. For a 500-person panel sending monthly waves with a small airtime reward, total monthly costs depend on country-level carrier rates and your incentive amount, both of which are estimable before you begin.
Ready to test before your first wave? Talk to Telerivet about setting up an account seeded with an example survey for your specific study design.