SBI Watchtower
A signals-driven application that correlates financial, market and investment data relevant to the U.S. Oil & Gas Rig Count.
Lead UX Designer | Hypergiant Industries | MVP Design Sprint | 16 Weeks
Overview
SB International, Inc. is a privately-held steel company located in Dallas, Texas. The company is a leading supplier and manufacturer of steel products for the oil and gas industry in The U.S. and Canada.
Their business process consisted of physically gathering data and other information from public data sources (ex. the stock market, financial and industrial news, social media, research, etc.) and mitigating as an executive team to determine business decisions and actions.
Problem
SBI executives were unable to see valuable data related to oil and gas rig production. They wanted a “crystal ball” to respond ahead of time to market demand for the steel parts they make for these rigs.
This data correlates with what is known as Rig Count, which is the total number of rigs in production as reported by the Baker Hughes U.S. Rig Count and the U.S. Energy Information Administration (USEIA).
Solution
SBI Watchtower is a web-based data management tool that collects, correlates, and relays information that enables users to make data-driven purchase decisions to maximize profitability and market share for their steel products.
It replaces outdated reactive triaging driven by revenue targets, and proactively surfaces production analytics, investment opportunities, and alerts users of risks.
Process
• Learn
• Discover
• Define
• Design
• Test
• Discover
• Define
• Design
• Test
Deliverables
• User Flows
• Info Architecture
• Task Flows
• Wireframes
• Interactions
• Info Architecture
• Task Flows
• Wireframes
• Interactions
Team
• Data Engineer
• Data Scientist
• Developer
• Senior Designer
• UI Designer
• Data Scientist
• Developer
• Senior Designer
• UI Designer
Tools
• Confluence
• Jira
• Figma
• Slack
• Google Docs
• Jira
• Figma
• Slack
• Google Docs
Research
Stakeholder Focus Groups
Focus group were conducted remotely via Zoom with SBI's President, VP of Procurement, Purchasing Coordinator, and Head of Customer Experience in Dallas, Texas. Hypergiant's VP of Design and Solutions Engineer were onsite to assist in facilitating, which included card sorting, a white boarding exercise, and completing a persona worksheet. High-level pain points included:
Focus group were conducted remotely via Zoom with SBI's President, VP of Procurement, Purchasing Coordinator, and Head of Customer Experience in Dallas, Texas. Hypergiant's VP of Design and Solutions Engineer were onsite to assist in facilitating, which included card sorting, a white boarding exercise, and completing a persona worksheet. High-level pain points included:
• Missing contextual data on their clients
• Unable to see financial data in real-time
• Unable to optimize their economies of scale
• Conducting business by instinct not information
Secondary Research
I compiled insights to identify key areas where increased transparency could benefit stakeholders. I reviewed information from industry reports, government publications, academic articles, and news outlets.
I compiled insights to identify key areas where increased transparency could benefit stakeholders. I reviewed information from industry reports, government publications, academic articles, and news outlets.
Technical Goals
1. Correlate open source data with internal KPI data
2. Generate and monitor prediction models
3. Create and disseminate analytics reports
User Stories
"As a user, I want to view U.S. Rig Count data and see what the count is predicted to be in the future so I can make informed business decisions."
"As a user, I want to view Rig Count data and other data related to the U.S. Rig Count so I can make informed business decisions."
"As a user, I want to generate a report based on the data I see so I can keep a record of it for reference in team meetings."
Feature Ideation
I facilitated three (3) one-hour remote workshops with members of the product team to share my research findings, and give them a pragmatic understanding of the business problem. Below is a screen capture of the Figma file for the first workshop.
Features
Data Correlation
Aggregated industry and market data related to Rig Count
Aggregated industry and market data related to Rig Count
Data Modeling
User-configured basic predictive data models
User-configured basic predictive data models
Data Reporting
Creating a record of a time-specific data visualization
Creating a record of a time-specific data visualization
Data KPIs
Comprising the top-level navigation, six (6) industry open-source datasets of key performance indicators (KPIs) were selected for building, training, and monitoring predictive data models:
1. Baker Hughes U.S. Rig Count
2. Standard and Poor's 500 (SP500)
3. West Texas Intermediate (WTI) Spot Price
4. Consumer Price Index - All Urban Consumers
5. Billion-Dollar Weather and Climate Disasters
6. Financial and Non-Financial News
Information Architecture
The final iteration of IA was reduced to open source data that correlated with the results of rig activity. Previous iterations included data related to drilling permits, which was considerably more granular and included data on drill activity, location, drill direction, production metrics and success rates.
Although this level of data transparency would have been valuable for the product, it was not readily accessible for use without permission.
Wireflows
The wireframe design process was speculative as I was worked in parallel with the Data Scientist and Engineer to determine which datasets were valuable for both the user and the data models. Below is the first wireframe/wireflow iteration where more than one rig count source dataset was an option.
Below is an example of the last iteration of the low-fidelity wireframe, and the high-fidelity wireflow for the Create Report feature.
1. Understanding the Supply Chain Business Model
Project Output
• Successfully aggregated open source data
• No granular data not accessible for modeling
• Never developed and prototyped
Assumptions + Observations
Assumptions: Data related to Rig Count would be available.
A lot of the rig production permit data was the property of private companies and not even provided for monthly reports by Baker Hughes or the USEIA. Rig Count data itself was too general to provide any real value for data analytics.
Observations: Rig Count data is unreliable for use in data modeling:
• Rig Count is rounded up to next whole numerical value
• A counted rig does not mean the rig is active
• Counts are lower in deep drilling areas
• No geological data is available
• Only reported monthly
Product Impact
In 2023, the features designed for Watchtower were repurposed for Hypergiant's Argus Command Center, an AI platform that provides accurate, real-time logistics data for:
• Command and Control
• Transportation + Logistics
• Energy and Load Management
• Physical Security
In Retrospect
From the beginning, this project had a tremendous amount of underlying ambiguity. Finding the value in the outcome was challenging. Once it was discovered that the necessary data was not available, it became impossible to complete the project.
A key takeaway from this project is:
Archive design work and create well curated artifacts. I do my best keep in mind that my work may end up in someone else’s hands. Having experience coming on to other projects after launch, I’ve learned to create artifacts with clear descriptions and explanations of my process.
