The Setup
Endava was hired by KIA Europe to run a discovery engagement for a new pricing tool. KIA's product pricing teams were managing the entire pricing operation through a collection of Excel files with heavy VBA automation. The goal was to understand the existing process, define the requirements, and design a modern web application to replace it.
I joined as UX Lead, working alongside a Business Analyst, a UX Designer and a backend Architect. I led the discovery discussions, mapped the process and designed the application structure and annotated wireframes.
The contract required complete Figma designs as well as discovery outputs. That meant taking an unfamiliar business process through to detailed application design within a schedule originally estimated for discovery.
The Mess
KIA Europe's pricing operation ran entirely through Excel. Not one file. A collection of interconnected spreadsheets with VBA macros, each one covering a different piece of the puzzle: ordering specifications, option codes, pricelists, P&L calculations, competitive analysis, CO2 taxation, FOB harmonisation with currency conversion.
One file alone, the Aggregation Master, had 48 sheets. It contained P&L breakdowns and pricelists for individual markets, each with their own option databases. A single option code grid could be over 200 rows by 270 columns. Then there were separate country-specific files, each with 30+ sheets of their own. On top of that, competitive analysis files pulling JATO industry data, CO2 taxation tables that varied by country, and ordering specifications with nearly 900 rows of configuration codes defining what could be combined with what.
Every day during discovery, another file would surface. We'd ask a question about one part of the process and the answer would be "oh, that lives in a different spreadsheet." Each file was another door to open and make sense of.
The domain itself was new territory. None of us had worked with an automotive company before. We didn't know what a "Trim" was, what "BlackTabs" meant, or how a car actually gets configured for different markets. It took several sessions before it clicked, and it clicked through a specific example: in Portugal, a base trim comes with standard seats. In Sweden, that same base trim might come with heated seats. Same car, same trim name, different contents depending on the country. That one example unlocked why the system was structured the way it was, and why it was so complex. The pricing wasn't just per car. It was per car, per trim, per market, with each market potentially redefining what a trim actually includes.
The three user groups reflected that complexity. Product Managers owned the ordering specifications, the master definition of how a car can be built. Pricing Managers worked across markets doing the actual pricing, P&L, and competitive analysis. NSC Market Users handled the country-specific side: local trims and taxation.
The spreadsheets gave each team a way to do its work, but the dependencies between files made the wider process difficult to follow. The design needed to preserve useful local flexibility while making shared information and decisions easier to manage.
Making Sense of It
We ran daily workshops with the Pricing Team, one to two hours each, where I led the discussions. The format was simple: ask questions about their process, look at the files together, map what we learned. Every session would surface another Excel file, another piece of the puzzle we hadn't seen yet.
It started making sense once I mapped the full process flow across the three teams. The Product Team owned the ordering specifications and the master configuration of each vehicle. The Pricing Team took that input and worked through Factory Trims, engine types, drivetrains, option combinations and exclusions, then the manual data entry phase covering CO2 values, FOB pricing, EU trim packages, and competitive JATO code matching. After a Kickoff, the Market Team for each country would create their local trims, build pricelists, run competitive comparisons, and submit for management approval.
Mapping the flow showed where responsibility moved between the Product, Pricing and Market teams. It also gave us a way to question steps inherited from Excel and decide which information belonged in shared configuration, which belonged in a pricing project, and when work needed to pass to another team.
| Pain point | Design response |
|---|---|
| Shared settings were embedded in the files used for daily pricing work. | Separate configuration from pricing projects |
| Local experiments could drift away from the central team’s official numbers. | Separate official pricing and simulations, with visible updates |
| There was no defined locking or approval flow for market simulations. | Make submission, review and revision explicit |
| An imported change could affect values elsewhere in the pricing workflow. | Specify comparisons, override priority and validation |
Creating Structure
Shared configuration
The first thing I did was separate configuration from workflow. In the Excel system, everything lived together. Currency exchange rates, taxation rules, country settings, model databases, JATO codes. All of it was embedded inside the same spreadsheets where people were doing their daily pricing work. If something changed at the configuration level, it rippled unpredictably through the files.
I pulled out six areas to manage independently: Currency, Countries and Markets, Taxation Rules, User Management, Model Management, and JATO Code Matching. In the design, these formed a shared configuration layer referenced by the pricing projects.
Taxation rules alone showed why this mattered. Austria had one rule, Belgium had four, and the Czech Republic needed separate calculations with and without CO2. France had a simulation rule alongside the current one. I brought these into a dedicated area with versioning, formula management and a record of who created or last modified each rule.
Official pricing and local simulations
On top of the configuration layer, I designed the core pricing workflow as a stepped process. Create a Pricing Project, select the car model and model year, upload the Ordering Specification, define packages. Then the flow split into two paths: Official pricing and Simulation. Both followed the same steps through Settings, Trim Configuration, Packages, Pricelist, P&L generation, and Summary, but they served different purposes.
This split addressed a tension in the Excel workflow. Market Users valued being able to detach from the central files and experiment, but the Pricing Team could not be sure that local numbers were accurate or current. I designed pricing projects that inherited official numbers as a starting point, with local overrides for simulations. When official numbers changed, the update would surface in the market projects so users could decide whether to adopt it.
Submission and approval
I designed the submission flow so a market simulation would lock while the Pricing Team reviewed it. The Pricing Team could run its own numbers alongside the submission without overwriting it. Approval would make the simulation official and feed it into the Aggregator; rejection would unlock it for revision and resubmission.
The proposed flow preserved the freedom to experiment while defining which numbers were official, when a submission could change and who reviewed it. These were interaction rules specified in the prototype, rather than outcomes measured in a live application.
Design and Specification
I did the wireframes in Balsamiq because it let me work fast and iterate during the workshops, sometimes adjusting flows based on questions that came up in the same session.
What happens when the data changes
The wireframes specified behaviour across the application as well as screen layouts. One example was importing updated FOB pricing from a model specification. FOB was a pricing input that could also be overridden at model or local market level.
- Show imported changes in blue, with the previous value available in a tooltip, so the Pricing Manager can compare before saving.
- Apply the documented priority: a local market FOB overrides a model-level overwrite, which overrides the base model FOB.
- Define file-upload validation, visibility and editing permissions, and what happens when someone leaves with unsaved changes.
I also documented edge cases inline, including a new currency year with no exchange rates entered. These annotations made the dependencies and exceptions explicit for the people taking the design forward.
I wrote the annotations for the UX Designer developing the high-fidelity UI and the Business Analyst writing the user stories. They connected the screen designs to the rules governing the wider application.
For the UI design, we used MUI (Material UI) because nearly every screen needed tables and inline editing. We aligned with the backend Architect on React as the proposed frontend framework and selected an external permissions tool tied to KIA’s Active Directory. The design system adapted MUI’s Figma components to KIA’s branding, keeping the effort focused on the interaction patterns.
The final Figma prototype covered the complete application: all configuration pages, the full pricing workflow for both Official and Simulation paths, the approval and locking flow, the PTR quarterly review process, aggregation reports, and the competitive analysis module. It was so large it had to be split across two Figma files.
The Outcome
We demonstrated the prototype to Market Users and gathered positive feedback on the proposed workflows. This was feedback on the design; we did not measure operational improvements in a deployed application.
The final deliverable was a fully specified Figma prototype covering the complete application across two files, along with the design system, information architecture, process maps, and annotated wireframes that could be used as a build reference. The stakeholder demo covering all flows ran just over an hour, which gives a sense of the application's scope.
KIA assigned implementation to another company. Our contribution ended with the design deliverables and handoff; this case study presents that work and the prototype feedback.