The Essential Engine

〰️

The Essential Engine 〰️

The Essential Engine

An internal data engine to find and launch new essentials at a rapid clip.


Role

Project Lead
Lead Designer

Year

2024

Disciplines

Human Interface Design
Interaction Design
Design Systems
Graphic Design

Introduction

The Engine scours and scrapes our competitors’ websites, Google search data and our own past product performance to understand exactly what is worth selling and how best to sell it.

It then supercharges our design and sourcing process for those products with insights about which manufacturers are behind the biggest brands and how much they charge.

Finally, it tells the story of the product by analyzing exactly what our users search for and which keywords they use when reviewing products.

It accomplishes all of this by using a host of AI and ML algorithms — resulting in a lighting fast and deadly accurate drop strategy.

I was the project lead and in charge of designing the software - meeting with team members and understanding their sourcing process, designing the user flow of interacting with the software, wire-framed the entire experience, and designed the interface.


Challenges

How can we design the experience so that the transition from one team member to the next is smooth?

How can we design the software to be a linear experience as well as jump between pages?

How can we integrate this software with AI tools to generate a detailed analytical report?



How the Engine Works

A detailed look at the engines launch strategy from beginning to end.

Finding What To Launch

We analyze a medley of first and third party data using a host of AI tools such as collaborative filtering to churn out meaningful and hidden insights.

The end result — a dynamic spreadsheet of the consumer products most likely to succeed on our site at any given point.

Creating The Perfect One

With spreadsheet in hand, our research team makes some decisions — which products would go best with our current catalog, what do we already know the supplier for, etc.

In doing this, our team is aided by the second part of the engine — a tool that algorithmically scours other retailers’ website and our own past data to understand what the best version of any given product is, how much it costs, what words customers use to describe it in reviews, how retailers describe it on their websites, which brands make it the best, etc.

The end result — we know exactly what The Essential version of a product needs to be.

Sourcing it

Knowing exactly what we need, we go out looking for it.

The engine scrapes all US customs and import data to tell us the names of the manufacturers that the big brands use, how much those manufacturers likely charge the brands and how much product the brand is even moving. The engine finds these answers using a host of AI and ML tools including neural nets and logistical regression.

Knowing this, we go straight to the manufacturers and ask them to make the same products with the same specifications for us — under our brand. Once they agree, we place the PO.

Launching it

Next, we write the pitch. This final section of the engine studies Google search data, the product listings on our competitors website and the reviews that customers leave on those listings. It churns out a set of keywords and phrases that are most commonly associated with a positive customer sentiment.

Our copywrited then writes email, push notification and product page copy centered around those keywords and phrases.


The Building Process

With each exploration, I focused on building prototypes to quickly learn from the team in their moment of need by providing iterations based on their usage & feedback.


The Essential Engine

A detailed look at the engines launch strategy from beginning to end.


Conclusion

The Essential data engine will allow us to launch new premium, private-label essential nearly every 24 hours.

By doing so, every 24 hours—

  1. We give non-members another reason to sign up In the form of a new digital ad and email campaign featuring the new product (created programmatically).

  2. We give members a reason to browse the store again In the form of a new email and push notification drop featuring the new product (created programmatically).

This will create our large, growing LTV and our low, falling CAC.

Please note, the UI and animations in this portfolio have been updated to reflect my current standards.