The most common marketing dashboard mistake isn't a wrong metric or a broken chart — it's building a single dashboard and showing it to everyone from the CMO to the campaign manager. An executive needs three or four numbers that answer "are we on track" in ten seconds; an operator running campaigns needs the granularity to answer "which specific campaign needs a bid change today." A dashboard built to satisfy both ends up too dense for the executive and too shallow for the operator — satisfying neither well.

Designing for the executive view

An executive dashboard should lead with business outcomes, not marketing activity metrics: revenue, pipeline, customer acquisition cost, blended ROAS — numbers that connect directly to a P&L line an executive already tracks. It should show trend against a goal or prior period at a glance (a single number with a direction indicator, not a 12-series line chart), and it should require zero interaction to understand — no filters to apply, no drill-down needed to get the headline. If an executive has to click into a chart to find out whether the quarter is on track, the dashboard has failed its primary job.

Designing for the operator view

An operator dashboard inverts most of those priorities: granularity matters more than brevity, campaign- and ad-set-level breakdowns are the point rather than clutter, and interactivity (date range controls, channel filters, drill-downs into underperforming segments) is a feature rather than something to hide. The operator dashboard's job is to surface where attention is needed right now — a campaign whose CPA jumped, a channel pacing behind budget — which means it needs more raw detail and more filtering surface than any executive view should carry.

Executive dashboard vs. operator dashboard

Dimension	Executive dashboard	Operator dashboard
Primary metrics	Revenue, pipeline, CAC, blended ROAS	Campaign CPA, CTR, spend pacing, CPM trends
Interaction required	None — headline visible immediately	Filters, date ranges, drill-downs expected
Update cadence	Weekly or monthly summary	Daily, sometimes hourly for active campaigns
Chart density	Low — 3-6 key numbers	High — granular breakdowns by campaign/ad set
Framing	Business-outcome language	Platform- and channel-native terminology
Failure mode if wrong audience	Executive loses interest, stops checking it	Operator can't diagnose issues, builds a shadow sheet

A vanity-metric dashboard is a design failure, not a data failure

Impressions, reach, and engagement counts are real numbers but rarely the right headline metric, because they can rise while business outcomes stay flat or decline — a campaign can generate huge impression volume with zero incremental revenue. A dashboard that leads with impressions or engagement isn't lying, but it's optimizing for what's easy to report rather than what matters, and it trains stakeholders to track the wrong thing. The fix isn't removing those metrics entirely — they're useful supporting context — it's refusing to let them be the headline number on any dashboard an executive will glance at for ten seconds.

Update cadence has to match the decision it supports

A dashboard refreshing hourly is wasted effort if the decisions it informs only happen weekly — and worse, a dashboard that updates less frequently than the decisions riding on it is actively dangerous, because someone will act on stale numbers without realizing it. The right cadence question isn't "how fresh can we make this" but "how often does someone actually make a decision based on this number," and the dashboard should refresh at least that often, with a visible staleness indicator (last-updated timestamp) whenever the refresh isn't continuous.

staleness_indicator_pattern.txt

text

Scorecard tile layout:

  +-------------------------------+
  |  BLENDED ROAS                 |
  |                                |
  |         3.4x      ▲ 0.3       |
  |                                |
  |  Last updated: 6:00 AM today   |
  |  Source: sessions_daily (BQ)   |
  +-------------------------------+

Every tile fed by a scheduled query or materialized view
should show when it last refreshed. A stale number without
a visible timestamp gets treated as current — and acted on
as if it were.

Staleness indicators matter more as pipelines get more automated

Ironically, the more automated a dashboard's data pipeline becomes, the easier it is for staleness to go unnoticed — a scheduled query that silently fails leaves a dashboard looking exactly as polished as one updating correctly, with no visual difference unless a last-updated timestamp or a data-freshness alert is built in deliberately. Treat a visible freshness indicator as a required element of the dashboard, not an optional nicety.

One dashboard, multiple audiences — the practical fix

The cleanest resolution isn't maintaining two entirely separate tools, it's structuring a single Looker Studio (or equivalent) report with an executive summary page as the landing view and a separate, more granular operator page linked from it — same underlying data source, different levels of abstraction. This keeps the data model consistent (no risk of the exec and operator views drifting apart because they pull from different queries) while still respecting that the two audiences need fundamentally different things from the same numbers.

What's next

Designing the right dashboard structure is only half the job — the calculated fields and formulas that make the underlying numbers correct (blended rates, weighted averages, period comparisons) are covered in detail in the Looker Studio Calculated Fields Cookbook.

Next: Looker Studio Calculated Fields Cookbook →

I build these systems professionally.

Whether it's a RAG pipeline, analytics migration, or AI workflow — let's talk.

Work With Me →

Need production AI or MarTech expertise? I build custom LLM workflows, server-side GTM tagging, and analytics infrastructures professionally.

Need custom AI or MarTech setup? Let's build together.

Let's talk →