Optimising global operations for shipping operators
Shell has been a pioneer in liquefied natural gas (LNG) for more than 50 years — making it one of the world’s largest LNG shipping operators, making up 20% of the global LNG shipping fleet.
Automating the operational processes presented the opportunity to optimise operators’ time and significantly reduce the possibility of human error through manual operations.
Service design, business analysis & product management as part of
Shell LNG Marketing & Trading’s Operations team.
Jun 2020 — December 2021
How might we improve global operational efficiency while reducing human error for one of the world's largest shipping operators?
The Role
Understand and define end-to-end user journey through user interviews and journey maps.
Work with Product Owner to prioritise backlog and user stories ahead of sprints.
Independently identify significant opportunities in optimising workflow and build consensus around roadmaps.
Moderate sessions with a full stack cross-functional product team to solve technical challenges.
Ensure product meets compliance standards for information risk management.
Test new functionality and ensure it meets specified user requirements prior to production release.
Build and maintain library of process documentation from design to build to release phases.
The Challenge
Shell’s LNG Marketing and Trading function trades up to 40 million tonnes of LNG per year, with global operations being managed by a team of 50-odd operators working across 4 time zones. For years, operators have been tracking their tasks, trade deal and voyage information, and commercial deadlines through manual management of Excel spreadsheets — leaving Shell exposed to human error, which can cost millions of dollars in levies and fines.
Automating a deeply integrated business unit presents multiple challenges, especially when there are many interdependencies within a typical workflow. Identifying the personas and day-to-day responsibilities was made inherently more difficult with the virtual-only constraint and end users working across four different timezones.
Full automation would also require integrations with third-party energy trading and risk management (ETRM) and industry-standard compliance software. The digital product solution would thus need to be delivered on the balance of desirability, viability, and feasibility with product, design, and tech teams working cross-functionally for successful delivery.
The Outcome
The release of the minimum viable product (MVP) and subsequent releases have resulted in:
83% of manually managed spreadsheets and trackers retired
6000 interactions automated annually, resulting in savings of ~500 full-time employee (FTE) hours
Increased visibility and monitoring on over 4,500 commercial deadlines and 43,500 standard tasks
Instant oversight into the 43,500 standard operational tasks managed annually
The Process
Work redacted for NDA.
01. User journey mapping
User journey maps were constructed following many hours of interviews and workshops which were instrumental to understanding the many interdependencies of systems, personas, and workflows within the business unit.
The process also informed the product features backlog based on high priority user requirements early on in the product build and delivery.
02. Collaborative ideation.
A regular cadence was set with users to ensure the ideas, feedback, and insights were continuously collected throughout the prototype build and test phase.
Collaborating with users in the ideation process also increased the level of trust in the digital product being built and encouraged users to be invested in the design process.
03. Prototyping with Figma
Following the refinement of requirements with users and the Product Owner, user stories were then created and wireframe prototypes were created in Figma to take out the guesswork from the developers’ end.
Prior to the use of lo-fi prototypes in the delivery cycle, developers would spend a much longer time figuring out the layout, placement, and functionality of requirements all through code-building.
Creation of low fidelity prototypes also allowed for quick validation with end users so that the feedback loop and turnaround time on iteration remained short.
04. Managing global users.
Since our users were split across three different timezones — Singapore, Dubai, and Houston — and the product team were split across London and India, the usability testing sessions had to be scheduled and managed carefully to ensure that the right combination of team members across the different time zones were able to attend the sessions.
Following each usability testing session, a 15 minute huddle was scheduled amongst the product team members to align on learnings, observations, and general takeaways that we could use to leverage on the next session.