A million emails a month

April was a milestone month for us. Goodlight AI is now sending over 1 million emails per month — each one hyper-personalised to the individual shopper receiving it. Not batch-and-blast. Not segmented campaigns. A million unique conversations, running autonomously every month. And as more shoppers join our clients' loyalty programs, that number will only grow.

The rest of the month was spent making the engine smarter: cleaner product data, better subscription controls, and a new approach to winning back shoppers who'd gone dormant. Read on for the details.

Win Back Campaigns for Hibernating Shoppers

Every retailer has shoppers who came in, bought a few times, and then disappeared. Reactivating them is tricky — you don't have much to go on, so personalization is hard and generic "we miss you" emails rarely work.

This month, we built out a dedicated reactivation campaign for these shoppers. The system identifies a top product pick for each shopper persona using our proprietary machine-leanring models and builds the email around that — specific enough to feel personal, without needing a full behavioral history to do it.

Integrated Channel Preferences

Shoppers who want to opt out of personalized marketing emails or texts can now do so directly through the Unsubscribe link in their emails — no need to dig into loyalty account settings. Think of it as a second layer of control that sits alongside the subscription settings on your loyalty page, making it easier for shoppers to manage their preferences and easier for you to stay on the right side of their inbox.

Smarter Product Data, Automatically

Product names in retail POS systems are messy — truncated, inconsistently capitalised, missing brand context. Getting them right matters because they show up directly in your shoppers' emails. This month, we automated the full pipeline that keeps our product catalogue clean and current.

Four workflows now run on a daily schedule: pulling in updated product images, enriching existing items with verified brand, category, and ingredient data, re-evaluating which products are actually marketable using AI, and automatically onboarding new PLUs as they appear in your POS sales data. The whole thing runs without anyone touching it.

—

Every retailer we talk to is sitting on years — sometimes decades — of inventory, sales, and purchase data. Most of it goes unused. We think that's a shame, because the path to real incremental growth is usually buried in there somewhere. If you'd like to dig in, we're always happy to chat.