Booth level management: what a panna pramukh structure really requires

One page of the electoral roll is roughly 30 voters. One page owner — a panna pramukh — who personally knows those thirty. For a 2.2 lakh voter constituency that is about 7,000 people. Above them, booth committees of 11–21 per polling station, and above that ward in-charges. Write the number down before you start, because most campaigns discover at T-30 that they have a third of what they assumed.

Take the party's booth list and call every name. In every seat we have audited, between 30% and 50% fail: moved away, no longer active, listed at two booths, or never existed. Do this in week one. A campaign plan built on an unverified list is a plan built on a number somebody made up at a district meeting in 2019.

A worker app with task assignment, completion proof and geo-stamped check-ins turns a list into an operation. Weekly task completion by booth is the single most predictive leading indicator we track — better than crowd size, better than social engagement. Booths where task completion falls below 60% for two consecutive weeks reliably underperform on polling day.

Unglamorous and decisive. Late or arbitrary payments to booth workers are the fastest way to lose a ground structure in the final fortnight, exactly when you need it. Route payments through the app against verified task completion, weekly, no exceptions. The campaigns that get outworked on polling day usually got outworked on payroll six weeks earlier.

Hourly turnout per booth from 7 AM, reported through the app or a WhatsApp bot. Compare against that booth's own historical pace, not against the constituency average. When a strong booth is 8% behind its own pace at 1 PM, move vehicles and callers there within the hour. That single mechanism has decided more than one seat we have worked.