The UX of Business Intelligence
Edge is a data driven web app delivering real-time visibility on key business metrics with artificial intelligence to optimise business operations and maximise returns.
Find out how Design Thinking led to a game-changing greenfield product for the petrol retail industry…
A Design Thinking Approach
Starting at the earliest discovery stage with a new start-up, the Palladium Product & Service Design team led a Design Thinking approach (empathise, define, prototype and test) to deliver a Minimal Viable Product (MVP) into the hands of petrol retail customers in less than six months.
The Team & My Role
I was the sole UX designer on an Agile Product and Service Design Team that comprised of a Service Design Director, Product Manager/Scrum Master and 4 developers. I was responsible for determining the overall design direction of the project, while collaborating with the rest of the team on all stages of the Design Thinking Process.
Designing the right thing
Empathise
The first stage (or mode) of the Design Thinking process involves developing a sense of empathy towards the people you are designing for, to gain insights into what they need, what they want, how they behave, feel, and think, and why they demonstrate such behaviors, feelings, and thoughts when interacting with products in a real-world setting.
The Challenge
There are approximately 8,718 Petrol Filling Stations (PFS’s) in the UK market. Early research segmented 6 distinct customer groups each with different characteristics.

1,809
Large Independent (100+ PFS’s)
479
Medium Independent (20-100 PFS’s)

1,307
Grocers
500
Oil Companies
4,624
Micro Independent (1-3 PFS’s) & Small Independent (4-19 PFS’s)
Methods / Activities
Empathy
Critical to our design process was the close partnership we formed with a small test group of users in the ‘Small Independent’ customer segment. This facilitated rich behavioural insights into the least supported group in all ‘modes’ of the Design Thinking process.
User Interviews & Workshops
Workshops with the start-up CEO, CTO and Product Manager were carried out to understand the business requirements in detail.
1:1 interviews with each of our 5 segmented customer groups of potential platform users were carried out.
Ethnography
Our friendly retailers, Raj, Tony and Jay were prepared to engage in ‘Day in the life’ on-site studies in the discovery, testing, prototype and MVP stages.
Journey Mapping
Understanding users journey’s and satisfying and unsatisfying experiences with current tools
Roles, Responsibilities & Goals
Regardless of the customer group, we discovered 4 key roles which could be combined by a single owner or held across an entire team in the case of larger retailers.

Owner
Responsibilities:
- Manages P&L and responsible for strategy
- Manages portfolio and high level site performance
- Review and authorise business decisions
- Understand the market and key business metrics
- Identify opportunities to grow the business
- Maximise volume, revenue and gross profit
- Manage and lead the team

Buyer
Responsibilities:
- Review, select and negotiate fuel supply contract
- Decide when to buy and at what price
- Manage current and future fuel levels at each PFS
- Ensure PFSs don’t run dry
- Ensure PFSs are on the best available fuel contracts
- Optimise buying behavior within contracts

Ops Manager
Responsibilities:
- Manage relationships with commission operators
- Facilities management, site security and compliance
- Financial reconciliations of tills to volume
- Keep the station running
- Reduce or eliminate station issues (in particular P1s)
- Review and authorise business decisions
- Ensure compliance and company policies are met at each PFS

Pricing Manager
Responsibilities:
- Pull together key pricing data from available systems and reports
- Manages portfolio and high level site performance
- Set the price at each PFS according to policy
- Set prices to maximise margin and / or volume in line with business goals
- Close the loop by ensuring sites follow agreed pricing
Define
Synthesise Observations
Identify/ Frame the Problem
The purpose of this stage of the process is to define the core problems (problem statements) from a human-centered point of view (POV) that you will tackle in the subsequent ideate phase. This is achieved by analysing and synthesising all the observations we have gathered in the Empathise phase.
Define Methods/Activities
Themes and Clustering
This is about making sense of the different data gathered in the research phase to reveal common themes, goals, pain points, patterns or concepts.
Personas
Personas provide meaningful archetypes that we use to assess our design against. They help us to ask the right questions and answer those questions in line with the user needs we are designing for.
P.O.V
Problem Statements
In Design Thinking, a Point Of View (POV) is a meaningful and actionable statement of the problem you seek to resolve.
Common Themes & Opportunity Areas
Through our user research we were able to identify pain points and common themes. Overlaid by the business objectives and a 5 year strategy, our minimal viable product focused on the following key themes:
Theme 1:
Adding financial value
Add financial value for petrol retailers by using data intelligence to optimise volume sales and margin
“I don’t have the know-how to reach my potential market”
“I don’t have the data I need to price accurately and with confidence”
“Can’t afford to take risks so I have to ‘top up’ whenever I might run low, even if the price is very high”
“It takes me 3 hours to calculate my margin across my portfolio of sites”
“Its hard to stay on top of managing risks and potential issues at every PFS”
Theme 2:
Improving productivity
Save retailers’ time with a convenient and intuitive platform that addresses many needs in a ‘one stop shop’
Theme 3:
Providing transparency
Build trust by providing light with an independent source of reliable, accurate and complete market and site information
“Our compliance tracking is all done offline. If there was a disaster we may lose the record of the safety precautions we’d taken”
“I don’t know whether sites are following the prices I have set”
“I don’t know what our other contract options are and its time consuming to shop around”
Personas
The Personas below show some of the fictional characters we created based upon our research, in order to represent the different user types that might use the product. Creating personas helps us to understand our users’ needs, experiences, behaviours and goals, keeping them front and centre of all design decisions.
Circumstance & Priorities
- Manages 11 sites for a mixture of fuel contracts across BP, Esso and Gulf
- Works closely with owner Raj
- Fuel is purchased by commission operators with guidance from Tony and Owner Raj
- Pricing decisions made by Raj and Tony
- Their focus is on the balance between margin and volume, aware that volume drives growth
- They are aware that most of their profit will be made on the property
- Attribute current business success to recent/current fuel prices
- Perception that location and brand are key customer differentiators, followed by service and price
Competitive Advantages
- Currently small enough to manage all sites without additional staff
- Location of sites
- Commission operator model
Key Goals
- Monitor margin and subsequent profitability
- Optimise pricing across portfolio to maintain volume and margin
- Pull together data/reports for review with Raj
- Make better buying decisions based on stock levels
- Negotiate better fuel contracts for the business

Key Responsibilities
- Daily monitoring of delivery volumes and sale price
- Make site by site price decisions based on performance metrics
- Distribute advice and information to sites on pricing/buying
- Ensure sites follow advice on pricing
- Support with resolving operational issues across the portfolio
- Due diligence on operations: review monthly safety and compliance logs
- Informing and advising sites on fuel orders for the day/week
- Ensure the site is “wet”interms of fuel availability whilst avoiding over supply
Key Challenges
- Understanding the best times to purchase fuel / pricing decisions – currently using multiple data sources to build this picture
- Gaining clear visibility of stock levels and ensuring they are ordering the right amount at the right time –particularly difficult for BP Smart Buy sites
- Understanding blended value of stock and therefore accurate margins
- Regarding fuel contracts, difficult to understand exact terms and make informed comparisons and decisions
- Reliance on multiple data sources –time and effort required to build picture of stock / margin / competitors
- If they added more sites, they would need another management layer
Circumstance & Priorities
- Own and operate 1 site each and have been in the petrol industry for years
- Current priority is volume, perceived as business growth
- Pricing is driven primarily by competitor activity (aim to price same or lower than their cheapest competitor)
- Manually calculate and set pricing for each PFS (review average barrel of oil cost, exchange rate, volumes of tanks)
- Perception that price is a key decision maker for customers, other considerations highlighted were brand, location, marketing and service e.g ATM, car wash, shop
Competitive Advantages
- Nimble and flexible, able to change price however and whenever they like
- They know understand their local customers –a sense of ‘giving back’ to their customers when margin is good, increases loyalty
- Local market knowledge and expertise
Key Goals
- Produce accurate reports in a timely manner
- Grow the business by driving volume
- Optimise pricing decisions to maintain volume and margin
- Decide when to buy and at what price
- Negotiate better fuel contracts

Key Responsibilities
- Pricing and purchasing decisions
- Site operations: staff, general maintenance, software
- Contractual negotiations
Key Challenges
- Time and effort – they are responsible for all business tasks, many of which are manual
- Customer perception that grocers are always cheapest
- Communicating/marketing to customers, they have little knowledge or capacity to do this currently
- Low margin / high turnover means that small mistakes in pricing can be costly
- Manually gathering competitor prices
- Data Fragmentation – difficulty in joining up sources of data to understand business performance
- Understanding potential new fuel contract pricing
- No visibility on actual Plattsprice
Designing things right
Ideate
Having gained a good understanding of our users from the Empathise stage, analysed and synthesised these observations into human-centered problem statements in our Define stage, we were now ready for the divergent phase of ideation.
How Might We Questions? (HMW)
How Might We?
Collaborative ideation workshops with the product team and our target personas were stimulated by “How might we?” questions based on our Problem Statements.
For example for our Brian the Buyer persona, we asked: “How might we… enable PFS buyers to buy fuel more strategically to optimise margin?”
Design Sprints
We estimated and prioritized our Product Backlog using DEEP principles. Our Product Manager defined acceptance criteria for testing and pumped User Stories into JIRA for bi-weekly design sprints (usually 2 weeks ahead of development).
Retrospective sessions
Daily retrospective sessions enabled the team to monitor progress and improve the approach for future sessions and/or design sprints based on individuals personal views and expectations.
Prototype
Rough sketching, task flows and wireframes progressed in fidelity as we gained confidence from continual testing.
Task Flows
Before diving into any prototype, I will iterate simple task flows to optimise the key user journeys and minimise friction. 37 Signal’s shorthand for User Flows is a great way to clarify and communicate ideas quickly with the team.

Rapid Prototyping
Early, low investment, black and white wireframes quickly came to life as interactive prototypes using Marvel. The Marvel App is a rapid prototyping tool that synchronises your Sketch workflow with an interactive cloud based prototype you can share and test with your users. It’s interactivity is basic but an invaluable tool for fast feedback in the early stages.
Test
Testing is undertaken throughout the process of a Design Thinking project. A range of methods were employed at different times, appropriate to our level of understanding of the problem.
User Acceptance Testing (UAT)
As a user story was developed through the process, from defining it’s acceptance criteria to possible solutions, I would share my designs or prototypes with the team in each sprint review for User Acceptance Testing. The client stakeholder (CTO), Product Manager and developer/QA’s in attendance would feedback to ensure each iteration met the acceptance criteria.
User Testing
Thinking Aloud
For me, the most effective and accessible user testing technique prior to development was the ‘Thinking Aloud’ usability test. Placing our friendly test users in front of one of my clickable prototypes, I would ask them to share their views on their overall feeling while using the app, usability, pain-points and areas where they were confused or frustrated.
When our users experienced difficulties, it was back to the drawing board (or Sketch) but with a streamlined workflow, iterations could be turned around quickly and tested again, often in the same day.
In some early iterations of the Performance Reporting UI, for example we had to accept the entire paradigm wasn’t working. In this case, I would then have to revisit our list of potential solutions and strategies to test new ways of solving the same problem.
Screen Recording & Heatmaps with Hotjar
Remote Testing
In both the latter testing stages of the MVP and post it’s official release, HotJar provided a unique window into our product in use on a daily basis.
This data that we obtained ‘in the wild’ was some of the most revealing and formed an invaluable bank of research for the subsequent Phase 1.0 release.
Living Style Guide
Throughout the UI design I produced a ‘Living Style Guide’ which evolved dynamically as design components advanced. Sharing the same symbols and styles between the page designs and style guide itself afforded agility and significant time saving.
I worked very closely with the Front End team to spec out any interactions that were not clear from the high fidelity mockups.
Impact
MVP Launch
After the MVP launched, we continued collecting stories from our users and gathering feedback. Comparing the stories we collected at the early stages of the project with those after launch, revealed the positive impact that our solution made on the lives of our Petrol retailers.
Clients who joined Edge before June 2018 have seen an average YOY profit increase of over 24% to June 2019.
+24%
YOY profit increase
EdgePetrol App v1.0
With a fully functioning product, the Edge team focussed on boarding new customers and connecting to a growing number of data points across the UK and Ireland.
Following the success of the MVP and a proven ROI for its customers, this time, Edge Petrol themselves invited me back to join the team for v1.0 of the app. Armed with three months of user data, new data points and a growing in-house team including data scientists things really got exciting!
Check out the Product Road MapA Start-up Success Story

US Expansion
Following expedited growth and work with major oil companies across the UK and Ireland, EdgePetrol has formed a close alliance with the industry-leading advisory company; Petroleum Equity Group (PEG), bringing EdgePetrol to retailers across the USA.
Read moreWhy Big Oil is BIG for EdgePetrol
EdgePetrol recently announced an exciting new partnership with Big Oil. As part of the partnership, EdgePetrol will now distribute Platts daily price assessments to Big Oil customers.
Read moreEdgePetrol goes live across 35 Certas Energy sites
Work across Certas Energy’s 35 UK sites to integrate Orbis Tech and EdgePetrol has been completed. The integration has facilitated the rollout of EdgePetrol to the Certas Energy sites, and marks the start of a close alliance between the two companies.
Read more