How to Build a Marketing Dashboard (Step by Step)

To build a marketing dashboard, start with the one question it has to answer, then work backward to the data. Pick the handful of metrics that connect spend to revenue (blended ROAS, blended CAC, MER, LTV:CAC, and revenue by channel), connect the sources that hold them (ad platforms, GA4, Shopify or Stripe, your email tool), reconcile the numbers against actual payments, and lay it all out on a single screen with a comparison period next to every tile. Everything else is decoration. A dashboard that takes forty minutes a week to maintain by hand will be abandoned by month three, so decide early whether you are building it, renting a BI stack, or buying something that blends the data for you.

How do you build a marketing dashboard?

You build it in six steps, in this order. Skipping the first step is why most dashboards die.

1. Decide the one question the dashboard must answer, and who reads it.
2. Pick the metrics that tie spend to revenue. Drop the vanity numbers.
3. Inventory and connect the data sources that actually hold those numbers.
4. Reconcile the data, because platform-reported conversions never tie to the bank.
5. Design the layout: one screen, top-line first, trend and comparison period on every tile.
6. Set refresh cadence and alerts, so the thing stays current without a human babysitting it.

Step 1: Decide the one question and who reads it

Every dashboard that gets used answers one question. "Are we spending profitably right now?" is a question. "Show me all our marketing data" is not, and it produces the 40-tile monstrosity nobody opens.

Write the question down before you touch a tool. Then name the reader. A CMO wants to know whether the quarter is on track. A paid media manager wants to know which campaign to pause today. Those are two different dashboards, and forcing them into one is the most common failure in marketing dashboard design. If two people would need different first tiles, build two views.

Step 2: Pick the metrics that tie spend to revenue

Once the question is fixed, the metric list mostly writes itself. The filter is simple: would a budget decision change if this number moved? Impressions, raw clicks, follower counts, and email open rates fail that test. Blended ROAS, blended CAC, MER, LTV:CAC, and revenue by channel pass it every time.

Cap the main view at roughly seven tiles. Seven is not a magic number, it is just the point past which people stop reading and start scrolling. Supporting detail belongs one click down, not on the front page. If you want the full list with formulas, we cover the fourteen marketing KPIs to track in a separate guide.

What metrics should be on a marketing dashboard?

It depends on who reads it. Executives need efficiency and payback. Channel managers need cost and conversion at the campaign level. Ecommerce teams live on AOV and repeat rate. SaaS teams live on payback and retention. Same underlying data, four different front pages.

Reader	Top tiles	Question it answers
Exec / CMO	MER, blended ROAS, blended CAC, LTV:CAC, total revenue	Is marketing profitable this month?
Channel manager	Spend, CPA, platform ROAS, CTR, conversion rate by campaign	What do I pause or scale today?
Ecommerce / DTC	Revenue, AOV, contribution margin, repeat purchase rate, MER	Are these orders actually making money?
SaaS	New MRR, blended CAC, CAC payback, trial-to-paid, LTV:CAC	Can we afford to keep spending?
Agency (client view)	Spend, revenue by channel, blended ROAS, week-over-week delta	Did we earn the retainer?

Step 3: Inventory and connect the data sources

List every metric from step 2 and write down exactly where the number lives. Most teams discover at this point that their single most important number, true revenue, lives in a system nobody had connected to the report.

Source	What it owns
Google Ads, Meta Ads, TikTok Ads	Spend, impressions, clicks, platform-reported conversions
GA4	Sessions, traffic source, on-site conversion rate
Search Console	Organic impressions, clicks, queries, position
Shopify	Orders, true revenue, AOV, repeat purchase rate
Stripe	Payments, MRR, refunds, churn, real cash collected
Klaviyo / HubSpot	Email and lifecycle revenue, leads, CRM stage

Connect these read-only. Nobody on the marketing team needs write access to Stripe to build a report, and a read-only connector is a much easier conversation with whoever owns security.

Step 4: Reconcile the data, because the platforms all take credit

This is the step that separates a real dashboard from a pretty one. Add up the conversions your ad platforms report and compare that total to the orders in Shopify or the payments in Stripe. The ad total will be higher, sometimes by 30 percent or more.

The reason is not fraud, it is overlap. A customer sees a Meta ad, searches your brand name, clicks a Google ad, and buys. Meta claims the conversion under its attribution window. Google claims it too. Both are telling the truth from where they sit, and if you sum them, you double count. Attribution windows differ (7-day click here, 28-day view there), modeled conversions get added on top, and none of it is audited against your bank account.

The fix is to anchor on money you actually received. Take total revenue from Shopify or Stripe as the denominator of truth, take total spend from every ad platform plus tools and agency fees, and compute blended numbers on top of those two anchors. Platform ROAS still has a job (it tells a channel manager which creative to kill), but it does not belong in the CFO's tile. Blended ROAS and MER cannot be inflated by overlapping claims, which is precisely why they belong at the top.

Step 5: Marketing dashboard design that survives contact with a CMO

Good dashboard design is mostly subtraction. A few rules that hold up:

• One screen, no scrolling. If the reader has to scroll, the layout has already failed to prioritize.
• Top-line first, top-left. Eyes land there. Put revenue and blended efficiency in that corner, not a traffic chart.
• Never show a naked number. Every tile needs a comparison: versus last period, versus target, or both. "MER 3.1x" means nothing. "MER 3.1x, target 3.5x, down from 3.4x" is an instruction.
• Roughly seven tiles. Detail goes on a second view, not squeezed into the first.
• One chart type per idea. Trend lines for time, bars for comparison between channels. Pie charts almost never earn their space.
• Label the source and the freshness. "Stripe, updated 09:14" ends the meeting argument about whether the number is real.

If you want a sense of what the finished thing looks like before you build, browse a few marketing dashboard examples and steal the layouts that put money at the top.

Step 6: Set the refresh cadence and the alerts

Match the refresh rate to the decision speed. Paid media decisions happen daily, so spend and CPA should refresh daily. Blended CAC and LTV:CAC are noisy day to day and are better read weekly. Board-level MER is a monthly number. Refreshing everything hourly just adds noise to metrics that need a bigger sample.

Then add two or three alerts so the dashboard works while nobody is looking at it. Useful ones: spend up more than 20 percent day over day, blended ROAS below break-even for three straight days, revenue from any channel dropping to zero (which usually means a broken connector or a dead pixel, not a dead channel).

Build vs buy: spreadsheet, Looker Studio, BI stack, or blended dashboard

There is no universally correct answer here, and anyone selling you one is selling you something. The honest comparison:

Option	Time to build	Ongoing cost	Right when
Google Sheets	A few hours	30 to 60 min of manual exports weekly	Two or three channels, one person, early stage
Looker Studio	1 to 3 days	Free tool, but connectors and blends break	Mostly Google data, client-facing reports
BI stack (warehouse plus Tableau or Power BI)	Weeks to months	Warehouse, ETL, and an analyst who owns it	Custom models, many sources, in-house data team
Blended dashboard product	Under an hour	Subscription, no maintenance	Blended ROAS, CAC, and LTV across paid, store, and billing

Should I build a marketing dashboard in Google Sheets?

Yes, if you run two or three channels and you are the only reader. A sheet with tabs for spend and revenue, a blended ROAS formula, and a weekly export habit will get you 80 percent of the value for zero dollars. It stops being the right call the moment a second person depends on it, or the moment the manual export gets skipped for two weeks, which it always does.

Is Looker Studio good for a marketing dashboard?

Looker Studio is genuinely good when your data is Google-native (Google Ads, GA4, Search Console) and you need something client-facing on a budget. It gets painful when you blend Shopify or Stripe with Meta and TikTok, because you end up hand-building the joins, the currency handling, and the deduplication logic. Those blends are also the parts that quietly break and hand you a chart that is confidently wrong.

When does buying win?

Buying wins when the maintenance cost is the real cost. If the numbers you actually need are blended (spend from four ad platforms against revenue from Shopify and Stripe), a purpose-built marketing dashboard software connects the sources once and computes the blended math for you, versus you owning the pipeline forever. MixedMetrics connects read-only to GA4, Google Ads, Meta Ads, TikTok Ads, Search Console, Shopify, Stripe, Klaviyo, and HubSpot, then puts blended ROAS, blended CAC, MER, LTV:CAC, and revenue by channel on one live marketing KPI dashboard, with an AI layer that flags what changed and where money is leaking. Pricing starts at $79 a month, which is cheaper than one afternoon a week of an analyst's time.

What to do once the dashboard tells you something

A dashboard is only worth the decisions it triggers. The most common thing a blended view reveals is that spend is not the problem. When traffic and spend hold steady but revenue slides, the leak is downstream: the conversion rate on the page people land on. That is the cue to audit the landing page copy, layout, and calls to action rather than reshuffle budget between channels that are all performing fine.

Other patterns worth acting on: blended CAC rising while platform CPA stays flat (you are buying the same customer twice), one channel taking credit for revenue that appears the day you pause it (attribution overlap), or MER holding while contribution margin falls (discounting is eating the win).

The short version

Pick one question. Pick the five to seven metrics that answer it and that tie spend to real revenue. Connect the sources, reconcile against the bank, put it on one screen with comparison periods, and refresh it on a schedule that matches how fast you actually decide. Then be honest with yourself about maintenance. A spreadsheet you keep current beats a beautiful dashboard you abandon, and a live blended view beats both.