Ask a business how marketing is performing and you will usually get several answers that do not reconcile. The ad platforms report one number of conversions, analytics reports another, and the finance system reports revenue that matches neither.
Nobody is lying. Each system is measuring something slightly different, defining it slightly differently, and none of them can see the others.
What is a growth intelligence layer?
A growth intelligence layer is a connected data and reporting system that sits above your operational tools, such as ad platforms, analytics, CRM and accounting. It joins their data on common keys, applies one agreed definition for each measure, and gives leadership a single trusted view of what drives growth, from marketing spend to revenue collected.
Is it a data problem or a connection problem?
The instinct is to collect more data. Almost always, the data already exists. What is missing is a shared definition and a place where the sources meet. The layer is that place, and it becomes the thing people actually look at.
What does a growth intelligence layer connect?
| Source | What it contributes | What it cannot tell you alone |
|---|---|---|
| Ad platforms (Google, Meta, LinkedIn) | Spend, impressions, clicks, platform-reported conversions | Which conversions became paying customers |
| Web analytics (GA4) | Behaviour, journeys, on-site conversions | Revenue that closes offline or later |
| CRM | Enquiries, pipeline stages, won and lost deals | What each lead cost to acquire |
| Accounting and business systems | Invoices, payments, fulfilment, retention | Which marketing produced the customer |
| Calls and WhatsApp | Conversations, sources, outcomes | Anything, unless the source is recorded |
The value appears at the joins. Spend alone tells you cost. Spend joined to closed revenue tells you whether the cost was worth paying, and that join is the one almost nobody has made. Our guide to marketing attribution explains how to carry campaign data through to the CRM so it can be made.
Why do definitions come before dashboards?
Most reporting disagreements are definition disagreements wearing a technical costume. Does a lead count when the form is submitted, or when it is qualified? Is revenue booked at order or at payment? Does a returning customer count as an acquisition?
Where does AI fit?
Once data is joined and defined, several things become possible that were not before, and this is where AI earns its place rather than decorating the pitch.
- Anomaly detection: flagging a channel degrading while there is still time to react.
- Segmentation on behaviour and value rather than on demographics someone guessed.
- Forecasting that accounts for seasonality instead of extrapolating last month.
- Scoring, such as predictive lead scoring, so sales calls the best leads first.
- Summarisation, so a weekly review takes ten minutes rather than an afternoon.
Each of these traces back to source data you can inspect. That matters: an insight nobody can verify is an assertion, and it will be treated as one the first time it contradicts somebody’s intuition.
Do you need a data warehouse?
Not on day one, and not to replace systems that work. Most mid-sized businesses can build a genuinely useful layer on the platforms they already run, using connectors and a reporting tool such as Looker Studio or Power BI, and only outgrow that once volume and the number of sources justify the step up to a warehouse such as BigQuery.
Anyone recommending a warehouse before understanding your volume is selling infrastructure rather than solving your problem.
What does building one involve?
- Diagnose: list the decisions leadership needs to make and the systems that hold the evidence.
- Define: agree one definition for each key measure, written down.
- Fix capture: tag campaigns, record lead sources in the CRM, track calls and WhatsApp. Our guide to connecting business systems covers the plumbing.
- Join: connect the sources and match records on common keys such as email, phone, order or deal ID.
- Report: build a small set of views that answer the decisions from step one.
- Extend: add alerts, forecasts and scoring once the numbers are trusted.
How do you know it worked?
The test is not whether the dashboard looks impressive. It is whether decisions changed, such as budget moved, a channel cut or a segment treated differently, and whether the people making those decisions stopped asking for the numbers to be checked first.
What does a growth intelligence layer cost?
It depends on how many systems are involved and how clean their data is. At Lulonke, the work starts with a diagnostic audit, and the initial architecture is typically R65,000 to R220,000. See AI growth intelligence and data and business intelligence for how we approach it.
Frequently asked questions
What is a growth intelligence layer?
A connected data and reporting system that joins your ad platforms, analytics, CRM and finance data, applies one agreed definition for each measure, and shows leadership a single trusted view of what drives growth.
Why do Google Ads, Google Analytics and my CRM show different numbers?
Each system measures something slightly different, with its own definitions, attribution settings and timing, and none can see the others. A growth intelligence layer reconciles them under agreed definitions.
Do I need a data warehouse for business intelligence?
Not at first. Most mid-sized businesses can start with connectors and a reporting tool such as Looker Studio or Power BI on their existing systems, and move to a warehouse such as BigQuery when volume and sources grow.
How is this different from a dashboard?
A dashboard displays numbers. A growth intelligence layer is the connected, defined data underneath that makes those numbers trustworthy and comparable across marketing, sales and finance.
How long does it take to build a growth intelligence layer?
It depends on the number of systems and the state of their data. A first useful version, answering a few key decisions, usually comes before the full build, and capability is extended from there.
