MANNA SNAP Intake System

The Problem

In June 2025, Feeding America ended a longstanding partnership with Google to funnel people interested in applying to SNAP to their local food banks. With this successful program now sunset, MANNA FoodBank needed a reliable, low-friction way to capture leads for SNAP assistance and immediately follow up across multiple channels, without requiring staff intervention at each step. Food-insecure individuals reaching out for help shouldn't have to wait for a manual response. Every hour of delay is a barrier.

The existing process was fragmented: form submissions landed in an inbox, follow-up was manual and inconsistent, and there was no way to track whether outreach converted to actual appointments.

The Solution

I designed and implemented a single-trigger automated intake system. A form submission at snap.mannafoodbank.org simultaneously fires outreach across three channels, logs the lead to a helpline outreach list, fires conversion tracking pixels, and routes everyone toward Acuity appointment booking, all without staff involvement.

Every architectural decision prioritized the person on the other end with a "white glove service" approach: immediate response, clear next steps, warm messaging. The system respects the urgency of someone asking for help.

The Hard Part Was Alignment, Not Automation

Wiring the integrations together was the straightforward part. The real work was human: getting several departments to agree on a single intake form. Every question on that form was a negotiation between three groups whose needs pulled in different directions.

MANNA's internal teams needed enough information to actually serve each person and to meet their reporting obligations, which argued for more fields. Digital marketing needed the form short, because every added question measurably lowers completion, and a form someone abandons helps no one. And the neighbors themselves, the people reaching out for food assistance, needed the lowest possible friction: someone asking for help should not be met with a wall of questions, and their dignity matters as much as their data.

I also negotiated with an outside vendor whose own requirements shaped which questions had to stay and which could be cut. My job was to hold all of these needs in tension at once and land on the shortest form that still carried the data the team genuinely relied on, removing everything that was only nice to have. That meant sitting with each department, understanding why every field mattered to them, and building consensus around a version that no single group would have designed on its own.

Workflow

Implementation Details