Date Toggles and Controlling Faceting

A date filter in the standard filter panel works, but it makes the user do work: open the panel, find the field, pick a range, apply. A toggle row across the top of the dashboard does the same job in one click — 2024 | 2025 | 2026 sitting there, obvious, always visible.

This lesson builds that row, and along the way covers the single most misunderstood setting in dashboard design: apply filters from faceting.

Every filter widget needs a step

Here's the rule that trips up nearly everyone the first time: a toggle, a list, or a date widget doesn't invent its own values. It reads them from a step — a saved query. If no step returns the field you want to filter on, the widget has nothing to display.

So the order of operations is always the same: build the step first, then drag the widget and bind it.

Group by a date bucket, not a raw date

You can't toggle on a raw date field — it has thousands of distinct values, and you'd get thousands of buttons. What you want is a date part: the year, the quarter, the month name.

CRM Analytics date fields automatically expose their component parts, and in a well-built dataset you'll often also have a custom bucket field (something like Fiscal Period or Close Year) created in the recipe precisely for this purpose. Group the step by that field and you get a handful of clean values — exactly one button each.

Filter the step so only relevant values show

Your dataset probably contains dates going back further than anyone cares about, plus a few stragglers from bad data entry. Left alone, your toggle row renders every one of them and wraps onto three lines.

Add a filter on the step itself — restrict to the last three years, or to a specific list of periods. The toggle only ever shows what the step returns, so filtering the step is how you curate the buttons. This is a query-level decision, not a display one, which is why it belongs on the step.

Cloning a step to make the second one

Once you've got your year toggle working, you almost always want a quarter toggle beside it. Building it from scratch means repeating every setting.

Instead, clone the step. Find it in the step list, choose clone, and you get an exact copy including the filters and grouping — then change just the grouping field to quarter, rename it something meaningful, and bind it to a second toggle widget. Thirty seconds instead of five minutes, and the two steps stay structurally identical, which makes them easier to maintain later.

The faceting gotcha

Now the important part. Drop your two toggles on, click 2025 on the year toggle, and watch the quarter toggle lose most of its options. Click a quarter and the year toggle may collapse to a single value. It feels broken.

It isn't. It's faceting doing exactly what faceting does.

Faceting means widgets on a dashboard automatically filter each other. Select a value in one, and every other step re-queries with that selection applied. For charts this is wonderful — that's the whole interactive-dashboard experience. For filter widgets it's often wrong, because a filter widget should show you all the choices available, not just the ones that survive your current selection.

The fix is a checkbox on the widget: apply filters from faceting. Uncheck it, and that widget's step stops listening to selections made elsewhere. Its buttons stay put no matter what the user clicks.

Do this for both toggles. Now the year row and the quarter row hold their shape, the user combines them freely, and the charts below respond to both. That's the behaviour people expect from a filter bar.

Tidying the spacing

Two toggle widgets side by side usually leave an awkward gap, because the dashboard grid puts space between cells by default. In the dashboard's layout properties you'll find cell spacing — drop it toward zero and the widgets sit flush against each other, reading as one continuous control strip rather than two separate boxes.

It's a small setting with a disproportionate effect on how "finished" a dashboard looks. Worth checking on every dashboard you build, alongside the surrounding page margins.

With filtering under control, the next thing to sharpen is how raw records get displayed. Next up: tables, images, and widget themes.