Marketing dashboards and reporting in Dubai

Most dashboards are a screenshot of the ad platform UI. They double-count conversions, ignore returns, can't show pipeline-graded performance, and break the first time a campaign structure changes.

1. 01

   1. Source-of-truth data model

   Pull from ad platforms, GA4, CRM and the data warehouse into a single model.

2. 02

   2. Reports designed for the audience

   Operator-grade weekly dashboards. Exec-grade monthly business reviews. Same data, different lens.

3. 03

   3. Maintenance plan

   Dashboards break when accounts change. We maintain them as a service.

• Source-of-truth data model
• Operator + executive dashboards
• Weekly reporting + monthly business review
• Dashboard maintenance and updates

Looker Studio, Power BI, or your warehouse + Metabase/Looker. We don't push a specific tool — we pick what fits your stack.

• Most UAE finance teams want unit economics, not spend totals. We design dashboards for the CFO.

Most dashboard arguments are definition arguments in disguise, so the first deliverable is a written metrics dictionary signed off by marketing and finance. It states what counts as a lead, a qualified lead, an opportunity and a sale, which conversion window applies, whether spend includes agency fees and VAT, whether revenue is gross, delivered or net of returns, and which currency the business reports in when ad accounts run in AED, USD and SAR. CAC is defined as spend divided by new customers, not all customers, and MER as total revenue divided by total media spend across all channels, so the two numbers cannot be gamed by moving budget between campaigns. For lead generation businesses the dictionary adds cost per qualified lead and cost per opportunity with the CRM as the source of truth. Each metric gets an owner, a source system and a refresh cadence. Only after this is agreed do we build anything, because a dashboard built on undefined metrics will be rebuilt within a quarter. The dictionary lives next to the dashboard and is updated when the business changes, with a dated change log.

The data model pulls from four kinds of source. Ad platforms, meaning Google Ads, Meta, TikTok, Snapchat, LinkedIn and DV360, come in through a connector such as Supermetrics, Funnel, Windsor or Fivetran, or through native connectors in Looker Studio for smaller setups. GA4 comes through its native connector or, for anything beyond basic reporting, the BigQuery export, which gives event-level data without sampling. The CRM, whether HubSpot, Zoho, Salesforce, Bitrix24 or Pipedrive, supplies leads, stages, values and dispositions with the click IDs and UTM fields captured on the lead record. The store, usually Shopify, WooCommerce or Magento, supplies orders, refunds, delivery status and product margin where it is maintained. Everything lands in a warehouse, normally BigQuery because it fits the Google stack and is inexpensive at this scale, where a small set of modelled tables join spend, sessions, leads and revenue on campaign, date and click ID. Refresh is daily for spend and sessions and hourly for leads where the CRM supports it. Access is granted to service accounts in your systems, so the pipelines keep running if we are not around.

The operator dashboard is built for the people who change campaigns. It opens on the week: spend pacing against budget by channel, cost per qualified lead or new-customer CAC by campaign, the tests running and their status, and a list of anomalies such as a campaign with spend but no conversions or a sudden CPM jump. Every chart can be filtered by channel, market, language and campaign, and every number links to the campaign it came from. The owner or finance view is monthly and fits on one screen: total spend, qualified leads or orders, pipeline or revenue, CAC, MER, contribution margin where it is available, and the trend over twelve months. It shows channel share of spend next to channel share of revenue so the allocation question answers itself. Both pages pull from the same modelled tables, so a number on the operator page always adds up to the number on the owner page. The monthly review is a written note on top of the owner page explaining what changed and why, because a chart without a sentence gets misread.

Google Ads, Meta and GA4 will each report a different number of conversions for the same week, and a dashboard that pretends otherwise loses trust the first time finance checks it against the bank. The differences have known causes: each platform claims the conversions it touched inside its own window, GA4 uses a different model again, the CRM only knows what sales recorded, and time zones and currencies add noise on top. Rather than pick one number and hide the rest, the model keeps all of them and labels them. Platform-reported, click-only and CRM-confirmed figures sit in adjacent columns, deduplicated figures are used for anything that rolls up to the owner page, and a reconciliation table at the bottom of the monthly view explains the gap channel by channel. Once the gap has been stable for a few months it becomes useful in itself: a sudden change in the ratio between platform and CRM conversions is usually the first sign that tracking has broken or that a campaign has started attracting junk. The rules below are applied in the model, not in people's heads.