A different way to think about scaling Platform Analytics
If every dashboard in your organization is "special," then none of them are.
There's a pattern I see across ServiceNow customers that quietly creates more problems than it solves: organizations treat dashboards as if they are analytics governance.
They're not. And confusing the two is what leads to the snowflake problem.
What's a snowflake dashboard?
The name earns itself. Like actual snowflakes, these dashboards are each unique — built from scratch for a single team, with KPI logic buried inside, owned by whoever built it, and slowly becoming something no one wants to touch. You've seen them. Maybe you've built them. (I have.)
They work until they don't. When something changes, you find out the hard way that the logic was never centralized. You fix it in one place and discover it lives in six others. Governance breaks down not because people didn't care, but because the dashboard became the governance layer by default.
What Platform Analytics governance is actually asking you to do
The Platform Analytics governance model has a clear priority sequence: govern your KPIs and analytics assets first. Make them trusted. Make them reusable. Then assemble dashboards from those assets, not the other way around.
When indicator definitions, thresholds, and breakdowns live outside dashboards, a few things get easier. Multiple dashboards can reference the same trusted data. Changes happen once. Governance becomes something you can actually maintain without a heroic effort every time something shifts.
The Platform Analytics Governance white paper lays this out in detail, and it's worth reading if your organization is still in the "who builds dashboards" conversation rather than the "who owns trusted insights" conversation.
The "prebuilt vs. custom" question is usually the wrong question
It's also a symptom of starting with dashboards. When the first instinct is "what do we need to build?", the conversation naturally lands on prebuilt versus custom. But that framing skips the layer that actually matters.
Prebuilt analytics assets in Platform Analytics are a starting point, a governed foundation. You extend where your operational model differs. You assemble dashboards to fit specific roles or workflows. You're not choosing between prebuilt and custom. You're combining them.
Platform Analytics ships with a meaningful library of prebuilt KPIs and dashboard templates tied to specific workflows — ITSM, CSM, HR, and others. For most implementations, a significant portion of what stakeholders ask for is already there. The real work is understanding what exists, validating it against your operational model, and identifying where you genuinely diverge. That assessment usually reveals less custom work than expected.
When something does look wrong — a metric that doesn't reflect your SLA commitments, for example — the instinct to build a custom KPI is usually the wrong move. The fix is more often upstream: update the SLA definitions in the operational layer, and analytics reflects that correctly. The principle is worth internalizing: analytics reflects operational truth, it doesn't replace it.
Where custom indicators are genuinely warranted, the discipline is keeping that logic in the KPI layer — owned deliberately, not buried inside individual dashboards where it can't be governed or reused.
Start with the questions that actually matter
When the entry point is "which dashboard should we build?", you're building from the wrong end.
Better starting questions: which KPIs do we actually trust? Who owns them, and is that documented? Where do different roles need to consume these insights?
Answer those, and dashboard design becomes almost straightforward. Skip them, and you're back to snowflakes.
Bringing it back around
Dashboards are a presentation layer, not a governance layer. They're how trusted insights get consumed, not where they should be defined.
Govern the KPIs. Reuse the assets. Build dashboards last.
Do that consistently, and something shifts: your dashboards stop being special, in the best possible way. They become reliable, maintainable, and easy to change. Which is exactly what governance is supposed to deliver.
For a full role-based breakdown of analytics governance in Platform Analytics, see: Governance in Platform Analytics experience
