2 min read
SMS: widest reach, lowest engagement
Post-DLT delivery runs 92–96% on clean data. It works on a feature phone with no data connection, which still matters in rural booths. But it is read-and-discard — treat it as a reminder and logistics channel (booth number, polling date, event time) rather than a persuasion channel. Regional-language Unicode cuts you to 70 characters per segment, so write to that limit rather than paying for four segments.
IVR voice: highest rural penetration
Answer rates of 38–52% in the 11 AM–1 PM and 6–8:30 PM windows. The candidate's own voice for 25 seconds outperforms almost anything digital in rural and semi-urban seats, because it reaches voters who do not read and do not use data. Add a press-key question and you get an 8–14% response rate — a rolling issue tracker at a fraction of survey cost, plus a warm contact list.
WhatsApp: lowest reach, highest persuasion
Reaches the smaller share of the electorate that is smartphone-and-data active, but with far deeper engagement and — critically — onward forwarding into networks you cannot buy. Local content forwards 6–8x more than party content. This is where opinion actually shifts, but it is not where you reach the whole roll.
Cost per meaningful contact
SMS is cheapest per delivery and most expensive per unit of persuasion. IVR sits in the middle on both. WhatsApp has the highest setup cost and the lowest cost per persuaded voter, but only after group infrastructure exists — which takes eight to ten weeks to build properly. That build time is why campaigns that start WhatsApp at T-45 never get value from it.
The sequence that works
Do not choose. Sequence. WhatsApp for persuasion from T-120 among the smartphone segment. IVR at T-90 and T-30 for the candidate's voice in rural booths. SMS at T-7 and polling morning for booth number and turnout. Three channels, three jobs, one message.
Bottom line
Channel arguments are usually vendor arguments. Pick by voter segment and campaign phase, and hold one message across all three.