B2B SaaS dashboard design, learned the hard way

When I joined PAFnow as founding product designer, the product was a dense ecosystem of dashboards — process views, event views, case views, analytical layers. Technically powerful, but overwhelming: data heavy screens, complex visuals with no clear path, and insights buried under layers of filtering.

The instinct in that situation is to simplify the product. That's usually wrong. The problem wasn't complexity — it was how complexity was presented. Raw data isn't understanding; clarity is. So instead of removing features, I focused on orchestrating information: what to show first, what next, and what can wait.

Every dashboard should answer three questions in order: what is happening, why it matters, and what to do about it. If a screen shows numbers without pointing at a next step, it's a report, not a tool.

At PAFnow, success meant a smooth flow from data to insight to decision. My rule for every design decision was simple: make the next step obvious. If a user has to think about where to click after reading a chart, the chart has failed them.

Enterprise users don't explore dashboards for fun. They arrive with a question. Structure each view around the single most important answer, and let everything else recede. At PAFnow that meant restructuring the experience so the process map prioritised critical information, the event filter guided users through intentional flows, and the Power BI dashboards replaced exploration fatigue with confidence.

A useful test: cover everything on the screen except the top-left region. If the user can still accomplish their main task, your hierarchy is right. If not, the important thing is buried.

At Heiland, a B2B e-commerce platform, the reorder flow was technically complete but took 18 minutes. Nobody lacked data — they lacked a path. Redesigning the dashboard around the reorder task cut that to 4 minutes, and 60% of dormant accounts came back.

Measure your dashboard by the time it takes a returning user to do the thing they came for. Every filter, tab and panel that doesn't serve that task is a cost you're charging your users on every visit.

In analytical products, colour is information. Use it consistently: one hue for the primary action, one for warnings, one for success states — and neutrals for everything else. When PAFnow got a cohesive visual language with colour used to carry meaning, it stopped feeling like a technical layer on top of Power BI and started feeling like a designed product.

Dashboards multiply. Without a foundation, every new view drifts. What kept PAFnow scalable was a design system with standard layout structures, typography and spacing rules, reusable components for charts, filters and navigation, and consistent interaction patterns across the platform. The system is what lets the product grow without the experience fragmenting — and it's a large part of what made the product credible enough to be acquired by Celonis.

The first session decides whether there is a second. PAFnow's guided experience — showing what to look at first, then what next — cut client onboarding cost by 80%. Don't treat onboarding as a tour bolted on afterwards; design the default empty and first-run states as carefully as the steady state.

Can a first-time user say what this screen is for in five seconds? Is the next action obvious from the primary view? Does the most important number survive a three-metre test? Is colour doing work, or just decorating? And can a returning user complete their core task in under a minute?

If any answer is no, that's where to start. If you'd like a second pair of eyes on your product's dashboards, let's talk — or see how these principles played out in the Heiland and PAFnow case studies.