Redesigning BP’s Global B2C Pricing Platform: From Legacy Input to Strategic Simulation
Back to works

bp Pricing Platform​

Client
bp

We replaced a fragmented 20-year-old legacy system with the unified B2C Pricing Platform. By integrating high-frequency Pricing, bulk GPC, and automated Rules into a single browser-based ecosystem, we transformed a high-stress data-entry task into a strategic advantage. The rollout was a definitive success, achieving rapid adoption in the DACH region and shifting analysts from reactive maintenance to proactive market control.

My role:
UX Designer
UI Designer
Prototype
User Research & Usability Testing
Design Systems

Platform:
Web

Duration
Jun. 2020 – Oct. 2023

Metrics

Measuring the work

0 %

Error reduction

0 %

Faster

0 %

Confidence rate

The context

BP’s pricing analysts are responsible for managing thousands of fuel stations in highly volatile markets. For two decades, they relied on a legacy .NET application that functioned like a static spreadsheet—offering no context, no safety rails, and no predictive capabilities.

The challenge

BP’s pricing analysts manage thousands of fuel stations in highly volatile markets. They were relying on a 20-year-old legacy .NET application that functioned like a static spreadsheet. It offered no context, no safety rails, and no predictive capabilities, forcing users to “fly blind” while managing millions of Euros in daily revenue.

The solution

I led the design of a comprehensive Pricing Command Center that transformed the workflow from manual data entry to strategic simulation. The new platform handles two distinct needs: Precision (managing specific station tactics) and Scale (executing bulk price changes across 500+ sites instantly).

Discovery and research

We started by interviewing Senior Pricing Analysts in Germany and Austria. We uncovered that the tool needed to support two completely different mental models.

The “Two-Speed” Market

Germany (The Speed Game):
The market follows “Edgeworth Cycles” (Sawtooth patterns). Prices crash and spike multiple times a day. Analysts need speed to react instantly.

Austria (The Strategy Game):
Regulations allow raising prices only once a day (at 12:00 PM). If an analyst misses the window, they are locked into a loss for 24 hours. They need prediction and safety.

The Trust Gap

Users ignored the old system’s data because they couldn’t distinguish between “stale” (old) data and “stable” prices. This forced them to waste time manually verifying competitors on external websites.

The solution

I led the design of a comprehensive Pricing Command Center that transformed the workflow from manual data entry to strategic simulation. The new platform handles two distinct needs: Precision (managing specific station tactics) and Scale (executing bulk price changes across 500+ sites instantly).

Personas

Claudia

Pricing Analyst (Austria), 36 years old.

 

Context:

Regulatory “hard stop” at 12:00 PM. High stakes.
Tech Literacy: Moderate. Transitioning from Excel; values clarity over speed.

Goal:

To set the optimal price for the next 24 hours. He needs to predict profitability before the 12:00 PM deadline locks him in.

Pain Points:

– “Fat-Finger” Anxiety: A single typo (e.g., 1.50 vs 1.60) creates an irreversible 24-hour loss.

– Flying Blind: She knows the price, but not the margin impact, forcing him to use external calculators.

Stefan

Pricing Analyst (Germany), 42 years old.

 

Context:

High-pressure environment, 3 monitors, continuously active.

Tech Literacy:

High (Power User). Prefers keyboard shortcuts and dense data.

Goal:

To react to market volatility instantly. He needs to match a competitor’s price spike in <60 seconds to capture the morning rush.

Pain Points:

– Click Fatigue: Performing 20+ bulk updates a day requires hundreds of clicks in the old tool.
– Naming Drudgery: Being forced to manually name every campaign wastes time.

The Result

The result was a shift from reactive data entry to strategic profitability, achieving a 100% reduction in critical errors.

100% Error Reduction:

Zero critical pricing errors (requiring database rollbacks) were recorded in the first quarter post-launch.

Methodology: Comparison of IT Incident Logs for “Emergency Reverts” pre-launch (Avg: 5/month) vs. post-launch (0).

Driver: “Diff-Aware” inputs and “Dry Run” simulation.

85% Faster Execution (Time on Task):

Reduced the time required for bulk price changes from 15 minutes to 2 minutes.

Context: Analysts previously toggled between 4 separate windows (Excel, Feeds, Legacy App) to manage 150+ sites. The new “Cluster Selection” and “Live Forecast” features allowed them to execute mass updates in a single click without context switching.

Metric: Time-on-Task study (n=5 analysts) validating the shift from manual site-by-site entry to bulk cluster management.

90% Confidence Rate:

Analysts reported significantly higher trust in their decisions due to real-time feedback loops.

Context: The legacy system provided no feedback on the impact of a price change until it was live. The new system displays immediate “Projected Margin” and “Competitor Gap” results before submission.

Metric: Post-launch survey (SUS & Confidence Rating) confirming that the “Live Forecast” feature directly reduced decision anxiety.

Design process

I began as the sole designer, eventually expanding the team to four designers working across five Agile squads. We adopted a rigorous iterative process, relying on continuous user research and usability testing to validate our decisions. We tested dozens of variations for every feature, refining the solution until it perfectly met the complex demands of the market.