Placing CPG brands on shelves at independent grocers
Co-built an AI-native wholesale service that helped CPG brands generate $17K in wholesale revenue across 120+ transactions.
I owned product and operations, client delivery, and ongoing development. I co-built the initial platform with a friend, then took ownership of the GitHub repository, retailer data, outreach processes, and automated reporting.
Results period: January–July 2026.
Emerging CPG brands wanted wholesale growth without the inventory commitments and cash demands of a national retail rollout. Independent stores offered another route, but reaching them required time that lean teams rarely had.
Even with a Faire storefront, brands still had to find suitable retailers, start conversations, coordinate samples, and follow up. Missed follow-ups and scattered information made it difficult to build a consistent sales process.
My bet was that what a store already stocked could help us identify promising prospects. We could then combine targeted matching with a managed outreach service to do more of the work for each brand.
What a store already stocked gave us a signal of potential fit. We compiled an internal database of 15K+ independent stores carrying 50+ emerging CPG brands, then enriched it with store type, food categories, location count, and stocked brands. This helped us prioritize relevant prospects for each client.
Using Claude Code and Python, we built a tool that takes a brand’s website URL and identifies relevant stores from our database. I then took over ongoing development, adding filters, improving the interface, and automating weekly client reports in Slack.
We were testing the AI-native service model outlined by YC: using software and AI to support delivery of the service.
We paired AI-assisted Instagram outreach with offshore operators who ran the day-to-day process. I developed SOPs and a CRM workflow to track conversations, samples, follow-ups, and purchases. Weekly reports gave clients a clear view of activity and results.
Instagram outreach averaged a 12% reply rate, compared with 1–2% in our previous email outreach.
The revenue figure reflects wholesale sales generated for client brands.
Optimize for the first order, not just the reply. Strong response rates helped us start conversations, but interest didn’t always translate into an immediate purchase. Shelf space, buying schedules, and seasonality shaped when stores could order. My next priority would be capturing those buying windows and using them to guide follow-ups.
The operating process matters alongside the tool. The retailer database, CRM, SOPs, and weekly reports worked together to keep outreach moving and make progress visible to clients. The service needed both useful software and consistent follow-through.
We used this database internally to match brands with relevant stores. Outreach was managed through a separate CRM we built in-house.
Enlarge screenshot