CASE FILE 02CLIENT GENERALISEDALL DATA MOCKED

Four ad platforms in one report.

An ecommerce agency ran ads on Meta, Google, TikTok and Bing. Each platform showed different numbers, and SEO was reported separately. This project put everything into one clear dashboard.

CLIENTECOM AGENCY
MODELPAID + ORGANIC ECOM
STACK4 AD APIS · GA4 · GSC · BQ · LOOKER
ENGAGEMENT5 PHASES · FULL SYSTEM
STATUSLIVE · AUTOMATED
THE BOARD — OVERVIEWAI-BUILT · LIVE HTML
EXHIBIT A
1

The main chart shows total revenue and total spend across all channels.

2

The bars show how much is spent on each channel. The funnel shows where visitors drop off.

3

SEO keywords appear on the same board, with their Google positions.

The Brief

Every platform showed a different number.

The agency ran ads on four platforms. Each one claimed credit for the same sales, so the totals never added up. Nobody knew the real return on ad spend.

SEO results sat in a separate report that was rarely opened. We moved all the data into BigQuery, calculated one combined number, and put paid ads and SEO on the same page.

THE GOAL: one report that shows all channels and SEO together, updated every day.

✗ THE PROBLEMS

  • Every platform claimed the same purchases — numbers never added up
  • Blended ROAS lived in a spreadsheet nobody refreshed
  • SEO reported separately, invisible on budget day
  • Cart → checkout drop-off unmeasured across sources
  • Budget shifts argued from four different dashboards
The System

How the system works.

Ad data from Meta, Google, TikTok and Bing is pulled into BigQuery every day, along with website and SEO data. SQL combines it into one set of numbers, and Looker Studio shows the results. Alerts run on top.

4 Ad Platform APIs

Meta, Google, TikTok, Bing — spend, clicks, conversions on schedule.

GA4 + Search Console

Site behaviour and organic queries — the unpaid half of the funnel.

Pipelines

Scheduled cloud functions land every source daily at 07:00 GST.

BigQuery

RAW → STAGING → MARTS. SQL dedupes cross-platform claims and defines blended ROAS once.

Looker Studio

Six pages: blended, per-channel, funnel, website, SEO, alerts.

Action Layer

Pacing alerts to Slack + a weekly channel-shift summary.

The Process

How the project ran.

01

Audit

  • Mapped duplicate purchase claims across all four platforms
  • Benchmarked GA4 vs platform-reported conversions
02

Measurement Plan

  • Blended ROAS + profitability defined once, in SQL
  • UTM convention enforced across every channel
03

Implementation

  • 4 ad APIs + GA4 + GSC → BigQuery, daily
  • Aggregated channel + funnel tables built and QA-ed
04

Dashboard Build

  • Six pages — paid and organic on one canvas
  • Funnel drop-off tracked per channel, LPV → purchase
05

Automate & Optimise

  • 07:00 refresh + Slack pacing alerts
  • Weekly channel-shift summary, auto-written
The Build

The channels, in detail.

Two more pages from the same dashboard: the channel table and the alerts page. Built in HTML with sample data.

PAGE 02 / 06 — CHANNELSLIVE HTML · MOCK DATA
EXHIBIT B
Channel Deep-Dive

Every channel, side by side.

  • The table shows spend, revenue, return and cost per sale for each channel — including organic.
  • The chart tracks the combined return against the 3.0 target line.
  • The bottom list shows which landing pages bring in the most sales.
PAGE 06 / 06 — ALERTS + AILIVE HTML · MOCK DATA
EXHIBIT C
Automation Layer

Alerts that come to you.

  • Rules watch the numbers and send a Slack message when something goes off track.
  • Every Monday, an AI summary explains the week in plain language.
  • Each alert includes a suggested next step.
Deliverables

What was delivered.

The Results

What changed.

ONE NUMBER.

All four platforms are combined into one daily return-on-spend figure that everyone can trust.

ADS AND SEO TOGETHER.

Paid and organic results now sit in the same report, so both are part of every decision.

A CLEAR FUNNEL.

The report shows exactly where visitors drop off, from landing page to purchase, for each channel.

NOTE: CLIENT DETAILS GENERALISED · ALL NUMBERS ARE SAMPLE DATA · DASHBOARDS ARE HTML RECREATIONS

Next Steps

Want this for your business?

If your ad platforms all show different numbers, this same setup can be built for your business.