Most PPC dashboards fail the same way. They are beautiful, they are comprehensive, they update automatically, and nobody can look at one and say whether last month made money. The problem is almost never the tooling. It is that the dashboard was designed around what the ad platforms export rather than around the decision it needs to support.
This is how to build one that answers the question. The metrics, the layout, the traps in platform-reported data, and where the whole exercise usually goes wrong.
Start with the decision, not the data
Before you open anything, write down the decisions this dashboard exists to support. For most paid media teams there are four:
Should we spend more or less in total? Which channel deserves the next dollar? Which campaigns should be paused? Is anything broken right now that we have not noticed?
That list is short on purpose, and it dictates everything below. A metric that does not change one of those four answers is decoration. Impression share is interesting. It has never once caused anyone to move budget by itself.
The metrics that belong on a PPC dashboard
Build in three tiers, top to bottom, and resist the urge to promote things.
| Tier | Metrics | Question it answers |
|---|---|---|
| 1. Money | Total ad spend, total revenue, blended ROAS, blended CAC, MER, contribution after ad spend | Did paid media make money this period? |
| 2. Channel | Spend, revenue, CPA, and conversion rate per platform, plus period-over-period change | Where should the next dollar go? |
| 3. Diagnostic | CPC, CTR, CVR, impression share, frequency, quality score, creative-level CPA | Why did tier 2 move? |
Nearly every dashboard I have seen inverts this, leading with a grid of CTR and CPC tiles because those are the fields the platforms hand over most readily. Those numbers explain performance. They do not measure it, and a stakeholder scanning the top of the page should not have to scroll to find out whether the account is profitable.
Tier 3 is still worth having, as long as it stays tier 3. Creative-level CPA in particular earns its place on paid social, because it is the metric that tells you a winning ad has fatigued and it is time to put fresh variants into the rotation rather than raise the budget on a tired one. Just do not let it sit above the money.
The double-counting trap
This is the single biggest reason PPC dashboards mislead, and it survives every redesign because it looks like arithmetic rather than a bug.
Suppose you spent $40,000 last month: $20,000 on Google Ads, $15,000 on Meta, $5,000 on TikTok. Google reports $70,000 in conversion value. Meta reports $55,000. TikTok reports $12,000. Sum: $137,000, for a reported 3.4 ROAS. Then you open Stripe and find $95,000 in actual revenue, which is a real ROAS of 2.4.
Nobody lied. A customer who saw a TikTok ad, later clicked a Meta ad, then searched your brand name and clicked a Google ad gets counted once by each platform under its own attribution window. The overlap is normally 30 to 60 percent of reported revenue, and it grows with the number of channels you run.
The fix is not a better attribution model. It is anchoring the top tier of the dashboard to revenue from your billing system or store, not to the sum of platform-reported conversion values. Platform figures still earn their place in tier 2 for relative comparison. They just do not get to define the total. Our breakdown of blended ROAS versus platform ROAS covers the calculation in detail.
Building it in a spreadsheet
A spreadsheet is a perfectly reasonable v1, and I would rather someone ship one this week than plan a platform for a quarter. The structure that works:
One tab per data source, each holding date, campaign, spend, clicks, conversions, and conversion value at campaign-daily granularity. One normalization tab that maps every platform's column names into a single schema, because Google calls it Cost, Meta calls it Amount spent, and TikTok calls it something else again. One revenue tab pulled from Stripe or Shopify. Then a summary tab that sums spend across all sources, pulls revenue from the billing tab, and computes blended ROAS, blended CAC, and MER.
Two things will break it. Platform exports change column names and date formats without warning, and manual refreshes stop happening in week three of a busy quarter. Both are survivable at one or two accounts. Neither is survivable across a portfolio.
Building it in Looker Studio
The free option, and a genuinely capable one. Google Ads and GA4 connect natively, Meta and TikTok need a third-party connector that usually costs money, and blending sources requires you to define join keys by hand. Expect a day of setup and a recurring maintenance tax when connectors expire or a data source schema shifts.
The honest assessment: Looker Studio is excellent at displaying data that is already clean and reconciled, and mediocre at the reconciliation itself. If your team has someone who enjoys this work, it is the best value in the category. If not, the dashboard silently rots and everyone quietly goes back to the platform UIs.
The layout that actually gets read
One screen. If the answer to "did this work" requires scrolling, it will not be read by the people whose budget decisions depend on it.
Top strip: four to six large numbers, all money. Spend, revenue, blended ROAS, blended CAC, and the change against the prior period. No charts, no gauges, just the figures at a size you can read from across a room during a standup.
Middle: one time-series of spend and revenue on the same axis, covering at least 90 days. This is where trend problems become visible before they become emergencies, and where a single chart beats a page of tables.
Below: the channel table, sorted by spend descending, with CPA and period change per row. Everything diagnostic goes on a second page or behind a click.
Alerts beat dashboards
The uncomfortable truth about dashboards is that they require someone to look at them. A dashboard nobody opened on the Tuesday a campaign's CPA tripled did not do its job, however good the layout was.
So pair the dashboard with a small number of threshold alerts, sent where the team already is. Blended CAC above target for three consecutive days. Spend on any campaign up more than 40 percent day over day. Revenue down while spend held flat. Four or five alerts, no more, or people learn to ignore them. This is the part that converts a reporting artifact into something that actually changes outcomes.
When to stop building and buy
The build-versus-buy line is genuinely about time, not money. Under three connected accounts with someone who enjoys the work, build it. Past that, the maintenance load compounds: more connectors to break, more schema changes to absorb, more reconciliation to redo when a platform revises attribution.
MixedMetrics exists for the past-that case. Read-only connectors to Google Ads, Meta, TikTok, GA4, Search Console, Shopify, Stripe, and Klaviyo or HubSpot, with blended ROAS, blended CAC, MER, and LTV computed for you and an AI layer that flags what changed and where spend is leaking, at a flat $79 per month. If you want to see the shape of it first, the Google Ads dashboard page shows the paid-search view, and ad spend tracking covers the multi-platform spend side.
Frequently asked questions
What should a PPC dashboard include?
A PPC dashboard should lead with total spend, total revenue, and blended ROAS across every paid channel, then break down to cost per acquisition, conversion rate, and CPC by campaign. Efficiency metrics like CTR and impression share belong lower, because they explain performance rather than measure it. If the top row does not answer whether the money worked, the layout is wrong.
How do you build a PPC dashboard in Google Sheets?
Export or pull campaign-level cost, clicks, conversions, and conversion value from each ad platform into one tab per source, normalize the date and campaign columns, then build a summary tab that sums spend and revenue across all sources and divides for blended ROAS. It works, but it breaks whenever a platform changes a column name, and it cannot reconcile double-counted conversions.
Why do my Google Ads and Meta numbers not add up?
Because each platform claims credit for the same conversion under its own attribution window. A customer who clicked a Meta ad on Monday and a Google search ad on Thursday is counted once by each. Adding the platform-reported revenue figures together typically overstates real revenue by 30 to 60 percent, which is why blended ROAS against actual billing data is the number to plan on.
How often should a PPC dashboard update?
Daily is right for spend and conversion data, which is frequent enough to catch a runaway campaign and infrequent enough that daily noise does not trigger bad decisions. Revenue reconciliation against billing can run daily too, but judge performance on rolling 7 and 28 day windows rather than single days, because most accounts do not have the volume for a single day to mean anything.
What is a good ROAS for a PPC dashboard to target?
It depends entirely on gross margin. At 70 percent margin, a blended ROAS of 1.5 breaks even and 3.0 is healthy. At 30 percent margin, break-even is around 3.3 and the same 3.0 is losing money. Set the target from your own margin and CAC payback tolerance rather than from a benchmark you read somewhere, and put that target line on the chart.
See how MixedMetrics works for your kind of team on the use cases page.
MIXEDMETRICS // GET STARTED
See this metric live across every channel
Connect your ad platforms, store, and billing through read-only connectors and watch blended ROAS, CAC, MER, LTV, and revenue by channel land in one live dashboard.