Skip to content

Campaign Suite: the dashboard for the whole team to see what each campaign brings in

My own measurement product, built for Kol Accesorios. It reads GA4, Search Console, Google Ads and Meta and shows the whole team, from the CEO to each area, what each campaign cost and what it brought, compared with similar campaigns.

Role
Product concept, measurement criteria and validation against the sources. AI-assisted development.
Year
2026
Stack
React 19TypeScriptViteNode / ExpressGA4Search ConsoleGoogle AdsMeta API (read-only)Gemini

Open the demo in a window

The idea

A business owner and their team need to decide how much to invest in ads by looking at a single screen. The suite brings the platforms together and applies its own measurement criteria, so every number is comparable and true.

  • A single result type per goal. Each campaign is measured by what it was after: reach, clicks, applications or sales.
  • Comparison between campaigns of the same type. Each one is seen against similar ones, with a verdict in plain words and an editable margin.
  • A common 28-day window, printed on every card and every table.
  • Three roles (view, edit, sync) and read-only access to the platforms.

Data checks

Before showing a dashboard, the suite checks that the numbers are true. In the real work it detected three measurement errors that became rules:

What it detectedWhat it does now
Meta contacts counted five times across platforms: 482 versus about 97 in the ad accountA single result type according to each campaign's goal
A Tag Manager tag in a loop: 1,531 events from 3 peopleDiscards events repeated hundreds of times per person
Meta at 90 days compared with Google at 28 on the same screenA common window, written on screen

The demo, tab by tab

Today, Campaigns, New campaign and Data checks. In each check, a switch shows the number with and without the error.

410 → 82
leads, with and without the error (demo)
1,204 → 8
“applications” from 4 people (demo)
$14,300 → $5,200
spend in the same window (demo)

Public demo with fictional data.

Get in touch