TL;DR

Enterprise decision-makers had powerful data.

They just couldn't act on it together, in real time, across departments.

The problem

Complex SAP analytics were being interpreted in silos. Collaborative, real-time decision-making (the kind that actually moves businesses) wasn't happening.

My role

One of two designers with equal ownership across the full experience: spatial UX, onboarding, boardroom, and business decision simulator. Embedded across three dev teams for six months.

The solution

A deployed VR prototype shown to enterprise clients and government delegations across three continents. Now being extended with Accenture, with Nvidia discussions underway.

Project context

What

VR Immersive Analytics Showcase

Client

SAP NVT & SAP Innovation Center

Timeline

6 months

Team

1 of 2 designers, 3 development teams

My Role

UX & UI Design
spatial UI
onboarding flow
wireframing
user journey mapping
prototyping
usability testing

Overview

SAP wanted to rethink how enterprise decision-makers engage with business data — not just view it, but step inside it. The Hybrid Business Steering Demo was their answer: an immersive industrial metaverse where SAP's analytics capabilities became a collaborative, spatial experience. The brief was ambitious and the medium was new territory. There were no established VR UX conventions to lean on, no design system to inherit. We were building the rules as we went.

The Design Challenge

What made this project genuinely difficult wasn't the complexity of SAP's data.

It was designing for a medium where the rules didn't exist yet.

No map to follow

VR has no settled UX conventions. No established patterns for spatial information hierarchy, gaze-based navigation or collaborative decision flows. Every decision had to be reasoned from scratch and tested with real users wearing headsets.

Five teams, one experience

Three development teams and two design teams with distinct responsibilities meant tight dependencies. Keeping design decisions coherent across the full pipeline meant constant structured communication and documentation.

Testing without a lab

You can't test a VR experience on a laptop. Every usability session required headsets, powerful hardware and a controlled environment. That constraint forced us to be deliberate about what we validated and when.

The process & design decisions

We couldn't design this like a web product.

So we didn't.

Due to a NDA, I can't share product screens beyond what's been publicly released. What I can share is how we thought.

Building a VR design framework from scratch

No guidelines existed, so we built our own. Legibility at distance, interaction feedback in 3D space, how to orient users without overwhelming them, etc.

Designing for collaborative decision-making.

This wasn't a single-user product. Multiple stakeholders would be in the virtual space simultaneously, looking at the same data and making decisions together. The layout, data hierarchy and interaction patterns all had to work from multiple angles.

Onboarding as a trust-building moment.

Early testing made one thing clear: Disorientation in the first 60 seconds killed engagement for the rest of the session. We rebuilt the onboarding flow to introduce the environment gradually. Controls first, context second, decisions last.

Simplifying without losing fidelity.

SAP's data is dense. The goal wasn't to summarize it. It was to make it spatially navigable. Made high-level KPIs visible at a glance. Details are revealed through deliberate interaction. The environment itself directs the attention, not instructional text.

The Experience Walkthrough

From first steps to final decision in four steps

01 Onboarding

Before anything else, the users get comfortable.

How to use the controls, navigation, how to move through the space. All happens at their own pace with no pressure to act. We treated this as the foundation everything else rested on.

Get it wrong and users spend the rest of the experience lost.
Get it right, and they forget they're wearing a headset.

02 The problem surfaces

Once users feel at home, the scenario begins - not through a briefing, but through discovery.

Mid-exploration, a notification appears: rising energy costs are threatening operational budgets. The users find the problem themselves. That distinction mattered to us.

Agency from the start.

03 The boardroom

Now that the problem is real, users bring their colleagues into the virtual boardroom to tackle it together. Live SAP data fills the space. Clicking highlighted areas reveals more detail and opens up possible actions.

The hardest design problem here was density.

The data needed to feel credible without being overwhelming, so we kept the main view clean and let users pull detail on demand.

04 The Business Decision Simulator

This is where the journey lands.

Users adjust key variables like energy prices, capacity, raw material costs and watch the impact update in real time. They can also move machinery around the factory floor, with every change feeding directly into the data.

Three scenarios to explore: Conservative, Realistic, Optimistic.

Once a strategy is chosen, users return to the boardroom to see how the numbers shifted.

The feedback loop between physical action and data response was the hardest thing to get right.

It was also the most important.

Impact & Reach

Deployed to enterprise clients and government delegations across 3 continents.

Strategic Partnership

SAP × Accenture

Hybrid business steering was named a flagship collaboration between SAP and Accenture at SAP Sapphire 2023 ↗

Potsdam, June 2023

Chenghua District Government

An urban development delegation from one of China's most active districts experienced the demo firsthand

São Paulo, June 2023

SAP Sapphire São Paulo

Presented at SAP's flagship Latin American enterprise event, drawing executives and customers from across the region ↗

Berlin, July 2023

SAP d-com

Showcased live to 50,000+ developers at SAP's flagship global developer conference

What I'd do differently today

01
Document decisions as they happen, not after. Across five teams and six months, too much design rationale lived in Slack threads and Figma comments. It slowed us down when we needed to revisit calls.

02
Test earlier, with less. Full VR hardware setup for every session meant our feedback loops were slower than they needed to be. A rougher proxy environment, used more often, would have caught things sooner.

New territory comes with new lessons. These were ours.

If you're still reading, we'd probably work well together.