Fast, AI-driven Insights for Shopify Merchants

Project: Midstream

Project context

Midstream is a Shopify app that enables users to harness AI. More precisely, Midstream states in the Shopify app store: “Brands collect vast amounts of data across inventory, sales, web traffic, and fulfillment, much of which remains underutilized. Only a small fraction of entrepreneurs know how to leverage their data, but with Midstream, that changes. Midstream allows any user to ask natural language questions of their data and receive precise, dependable answers. No dashboards, downloads, or training — just human answers to human questions.”

Working on this project excited me because I found the concept to be compelling. I have been on the analytical side of large amounts of customer data and it can be vast and challenging to quickly navigate. Therefore, the prospect of having the ability to query that data struck me as both practical and impactful. And it just seemed like a fun undertaking.

Credit where credit is due:

• One project manager
• One UI/UX designer
• The Midstream founders

My role & contributions:

• User experience design
• Visual design
• Interaction design

The challenge

This project was brief and one that I worked on for only a few days in between other assignments. Upon joining, I found that a preliminary set of wireframes had already been developed. However, I was asked to help further realize the user interface and user experience by incorporating visual design and branding, and delivering a functional prototype with interactions for beta testing.

The primary goal was to create an MVP.

The provided wireframes

Wireframe for the login page that allows users to integrate their Shopify storefront with Midstream.

Wireframes outlining both states of the Midstream query experience: on the left, the initial or 'empty' state before questions are asked; on the right, the state in which answers are returned.

The revised “answer” state, showing emphasis on the answer, related questions, and upcoming product features.

Design direction

This application is primarily an analytics tool, yet the functionality and user interface align with the way a search engine receives and presents information. To enhance the user experience, rigorous competitive research was not needed, but I aimed to keep the overall experience aligned with Google’s search experience. Compared to other search engines like Bing and Yahoo, Google takes a very lean approach to information disclosure. During the process, I kept these main characteristics top of mind:

  • Aim for a minimalist design, emphasizing the search bar and result presentation.

  • Provide visually clear and prominent answers, free from distracting graphical elements.

  • Expose primary information and conceal secondary/supplemental information.

In terms of branding and executing the visual design, the color palette and fonts were provided, but there were no strict rules regarding their implementation. Aiming once again for a minimalist approach, I used two colors most prominently, with the rest of the palette serving smaller moments like CTAs, to highlight specific headlines, the loading screen, and the mobile menu.

The final design

Finalized login page for users, showcasing current platform integrations and future platform integration options.

The finalized “home” state for asking questions with question recommendations.

Loading screen while results are being gathered.

The finalized results page, showcasing the answer, logic behind the answer, follow-up questions, and upcoming product insights.

The mobile experience from login to question answered.

A demo showcasing user interactions.

Impact & takeaways

This project was a quick but rewarding experience, enhancing my skills in interactive design within a fast-paced environment. Despite the brief development period, it was a joy to tackle the creative challenges.

Key takeaways include the importance of rapid execution and the realization that a product's success extends beyond design, highlighting the need for continued marketing and support post-launch.

While the product has not seen extensive use since its release, the experience gained remains invaluable. I stand by what I provided our client and I had fun doing it.