Designed a new feature that made 32% more contacts reachable by automations, delivered with AI in 6 days

Stores pay edrone for their whole contact base, but automations, which bring half the revenue they get from edrone, reached only 3.1% of it.

An automation can only reach a contact identified by a cookie, and nothing in the product helped a store raise that. Stores saw little return for what they paid, and churn was rising.

To understand what drives identification and how stores see it, I went five ways:

• Ran 6 user interviews to test my hypothesis that most stores did not know identification existed
• Ran internal research with the Support Team on how they were already dealing with it
• Checked the data in Tableau and compared store cohorts to see what affects identification
• Reviewed whether and how competitors handle identification
• Researched technical ways to raise identification and talked them through with developers

Support was already raising identification by emailing a store's whole base, and the research pointed the same way, since more opened emails mean more identified contacts. I turned that into Contacts activation, the first step of a larger area for activating a store's base, built around an automated email sequence:

• Branded as the store and sent automatically, with every email generated by AI.
• Multiple different emails, so a monthly send keeps getting opened.
• A 30-day cycle, because cookies clear on their own within about 30 days.
• Neutral content, not marketing, so it can go to the whole base.

On the dashboard it is a module with a preview of every email, its sending details and its performance. The interviews showed some stores did not get identification, so it also has a short How it works section.

Most importantly, it is on by default, shown in onboarding as already running, since stores would never look for a feature they did not understand.

With one backend developer and Codex, we shipped the feature in 6 days.