Loading tutorials…
Loading tutorials…
Filters and controls are what turn a static report into a self-serve dashboard. Get the scoping right and viewers answer their own questions; get it wrong and you'll spend Mondays answering 'can you filter this to Q1?' emails forever.
Who this is forAnyone whose dashboard is read by more than one person. If multiple stakeholders ask similar 'can you slice this by X?' questions, filters and controls are the cure.
What you'll need
Step 1
Looker Studio has four main controls: Date range, Drop-down list, Fixed-size list, Input box, Advanced (with date comparison). Pick by interaction pattern.
Date range control — picker for date scope. Add to every page that has time-based data. Set default to Last 28 days for marketing dashboards, Last 30 days for finance.
Drop-down list — single or multi-select from a dimension's values. Use for Campaign, Source, Region. Multi-select is the default — viewers can pick multiple values.
Fixed-size list — like drop-down but always visible (not collapsed). Use for short lists (3-5 options) where you want viewers to see all choices.
Input box — free-text search/filter. Use for unbounded dimensions (Search term, Page URL). Less common in marketing dashboards.
Advanced filter — date comparison (vs previous period, vs previous year). Add to executive overviews.
Step 2
A control can scope to one chart, one page, or the entire report. The right scope depends on what viewers expect to control.
Page-level (default) — control affects all charts on the current page. Use for most date range and dropdown controls. Drop the control on the page and it works.
Report-level — control affects every page in the report. Right-click the control → Make report-level. Use for filters that should be sticky across navigation (e.g., a Region filter that applies everywhere).
Chart-level — actually a filter, not a control. Charts inherit page/report filters by default. Apply chart-level filters via the chart's Filter property to scope tighter.
Recommendation: report-level for date range and primary segmentation (region, brand, account). Page-level for context-specific filters (search terms only on the search-terms page).
Tradeoff: report-level controls always show on every page (takes screen real estate). Page-level only appears where added.
Step 3
Every page with time data needs a date range control. Set default to Last 28 days. Make it report-level if the dashboard is mostly time-based.
On the Overview page, Insert → Date range control. Drop it at the top, below the title.
In the right panel → Properties → Default date range. Pick Last 28 days (calendar days, not weeks).
Decide on report-level vs page-level. For most marketing dashboards: right-click → Make report-level. Now every page inherits the same date scope.
Style the control: small font (10-12px), no border, neutral background. The control should feel like a setting, not a chart.
Test: change the date range to Last 7 days. Every chart on the page should update within 2-3 seconds. If a chart doesn't respond, it has a hardcoded date filter overriding the control.
Step 4
For each dimension viewers will commonly filter by (Campaign, Source, Channel), add a Drop-down list control.
Insert → Drop-down list. In the right panel → Data tab → Dimension = Campaign name (or whatever segmentation matters most).
Style tab — set "Show search box" to ON if there are many values. Set "Allow multiple selections" to ON (default).
Position drop-downs in a row at the top of the page, after the date range. 3-5 controls is the comfort limit; beyond, viewers get confused.
Always provide an "All" option in the data: most drop-downs default to "(All)" at the top, but verify it's present.
For long lists, set a sort order (Style tab → Sort): typically alphabetical by dimension value, or descending by a metric like Cost.
Step 5
Some charts should always show a fixed subset (e.g., "Brand campaigns only"). Apply chart-level filters that the viewer cannot override.
Click the chart you want to fix-scope. In the right panel → Filter → Add a filter.
Configure: Include / Campaign name / Contains / 'Brand'. The chart now only shows campaigns whose name contains 'Brand'.
Chart-level filters compound with page-level controls. If the page Date range is Last 28 days AND the chart filter is Brand campaigns, the chart shows Brand campaigns over the last 28 days.
Use chart-level filters for: focused-comparison charts (Brand vs Non-brand), funnel-stage charts (top-of-funnel only), and exclusion charts (everything EXCEPT internal traffic).
Document chart-level filters: rename the chart to make the filter obvious. 'Sessions — Brand Campaigns' is better than 'Sessions'.
Step 6
Filters can interact unexpectedly. Test: pick a value in one control, navigate to another page, verify the expected behavior.
Set the report-level Date range to Last 7 days. Pick "Acme Brand" in the Campaign drop-down. Navigate to Page 2.
Verify: Date range is still Last 7 days (report-level inherits). Campaign filter — if it was page-level on Page 1, Page 2 is unaffected. If it was report-level, Page 2 is also filtered to Acme Brand.
Test reset: click "Reset" (top-right of the control) — does it clear cleanly? Some controls have edge cases where reset leaves stale state.
Test with multiple selections: pick Brand + Non-brand in the Campaign filter. Does the page show both?
Document the filter model in a Methodology section if it gets complex. Most dashboards only need a 2-3 line note.
Step 7
Looker Studio supports click-to-filter on charts and drill-down dimensions. These are the next level of interactivity beyond explicit controls.
Cross-filter — when a viewer clicks a row in one chart, other charts on the page filter to that value. Enable per-chart: right panel → Interactions → Apply filter (ON) AND Cross-filter (ON).
Best for: a top-N table where clicking a campaign filters the trend chart below. Viewers love this — it's the difference between 'reading a report' and 'exploring data.'
Drill-down — add multiple dimensions to a chart and let viewers click to expand. Configure: chart → Data → Drill down → toggle ON → add 2-3 levels (e.g., Country → Region → City).
Bookmarks — viewers can save filter combinations as bookmarks. Encourage stakeholders to bookmark their favorite views — it reduces "can you save this filter as default?" requests.
Test all interactivity in View mode (not Edit mode). Interactions behave differently between modes.
Common mistakes
Adding 10 controls and overwhelming viewers
What goes wrong: A dashboard with 10 filter controls looks like an airplane cockpit. Viewers skip the controls and ask you to filter for them — defeating the purpose. The hours invested in controls produce zero engagement lift.
How to avoid: 3-5 controls per page max. Pick the dimensions viewers actually slice by, not every possible dimension. If unsure, ask 3 stakeholders "what would you want to filter by?" — pick the top 3 answers.
Page-level filters where report-level is expected
What goes wrong: A viewer picks 'Region: West' on Page 1, navigates to Page 2, and the filter disappears. They get a different (wider) view than expected, make a decision on the wrong number, and lose trust.
How to avoid: For primary segmentations that should persist across pages, right-click → Make report-level. For page-specific filters, leave page-level.
Hardcoded date filters in individual charts
What goes wrong: A chart has a hardcoded 'Last 30 days' filter. The page-level Date range is 'Last 7 days'. Viewers see the page set to 7 days but the chart shows 30 days — and don't know why.
How to avoid: Set chart-level Date range to "Auto" (inherits from page control). Apply explicit chart-level date filters only when you want to override.
Filters on dimensions the chart doesn't use
What goes wrong: You add a Campaign filter. One chart on the page has no Campaign dimension. The filter does nothing to that chart — but viewers expect it to. They see numbers that don't change when filtering and don't know if it's a bug or intent.
How to avoid: Verify every control affects every chart on the page. If a control should not affect certain charts, exclude those charts via Interactions → Apply filter → OFF. Document the exclusions.
Not testing in View mode
What goes wrong: Interactions work in Edit mode (where you build) but fail or behave differently in View mode (where viewers see). You ship a 'working' dashboard that's broken from the viewer's perspective. Embarrassment + rebuild cost.
How to avoid: After every meaningful change, switch to View mode and test interactions. Especially cross-filter and drill-down — they're Edit-mode-friendly but View-mode-finicky.
Recap
Done — what's next
How to set up Looker Studio and build your first report
Read the next tutorial
Hand it off
Interactive dashboards are how Looker Studio earns its keep. The control architecture takes practice to get right — and the wrong architecture creates more support work than it saves. A specialist can build a 3-page interactive dashboard with proper control scoping in one afternoon, typically $150-300.
See specialist rates
A control is a viewer-facing UI element (dropdown, date picker, search box) that lets the viewer filter charts. A filter is the underlying logic applied to a chart. Controls produce filters. You can also add filters directly to charts that viewers can't change.
Technically unlimited. Practically, 3-5 per page max. Beyond that, viewers don't use them. Add a Methodology section explaining what each control does if you need more than 5.
Sort of — viewers can bookmark a filtered view via Share → 'Copy link' (the URL captures filter state). They cannot save filter state to change the default for everyone — only the report editor can change defaults.
Three usual causes: (1) the chart doesn't have the filtered dimension in its data, (2) Interactions → Apply filter is OFF on that chart, (3) the chart uses a different data source than the control. Open the chart's data tab and check.
Looker Studio
Looker Studio is free and surprisingly powerful — but the first report is where most owners stall. This is the walk-through that gets you from blank canvas to a shareable dashboard without making the rookie structure mistakes you'll regret later.
Looker Studio
The native GA4 connector looks like three clicks — but the choices you make in those clicks lock in sampling, scope, and refresh patterns for the life of the report. Here's the right way to wire it up.
Looker Studio
Calculated fields are how Looker Studio goes from 'shows data' to 'answers questions.' The syntax is its own dialect — not SQL, not spreadsheet. This tutorial covers the formula patterns that pay for themselves in week one.
Looker Studio
The Looker Studio template gallery saves you 6+ hours per dashboard. A custom brand theme adds another 2 hours of consistency. Combined, they turn 'every dashboard is a snowflake' into 'every dashboard launches in 30 minutes.'