A marketing report should help decide where to act. For an agency, that requires relating traffic, enquiries, conversations and commercial outcomes without confusing them. This guide aims to build a weekly view showing where losses appear in the journey and which information is missing to explain them.
Start by writing a specific question: “Which campaigns generate enquiries that reach a diagnostic?” That question determines the data needed. If you collect indicators without deciding what to compare, you may end up with an extensive dashboard that guides no action.
Define a metric dictionary#
Agree on the meaning of every event: valid request, booked meeting, completed meeting, proposal and won opportunity. Add an owner, data source and recording time. A valid request might exclude tests, duplicates and submissions without an identifiable need, but the criterion must remain consistent throughout the comparison.
HighLevel maintains first and latest attribution information for contacts and records attribution when supported interactions occur, such as form submissions or bookings. Latest attribution can change with newly recorded interactions. Check which events and tracking mechanisms your implementation covers. Official documentation.
Define a naming convention for campaign parameters and maintain an inventory of links. Avoid using different names for the same source. If traffic is unclassified, investigate the journey and configuration before manually assigning it to whichever channel seems most likely.
Build a five-number view#
- Relevant visits to the acquisition journey.
- Valid requests received.
- Completed meetings.
- Proposals presented.
- Won opportunities and value recorded using a consistent criterion.
Add media spend when available and clearly associated with the period and campaign. Keep denominators visible: meeting rate per request and proposal rate per meeting answer different questions. If commercial data is updated late, display that limitation with the report.
HighLevel provides page statistics and a broader website and funnel analytics view. Figures can differ from other tools because of definitions and tracking mechanisms; do not force artificial agreement without reviewing how each system counts. Official documentation.
Interpret a complete cohort#
Hypothetical example: one campaign produces one hundred enquiries and ten meetings, while another produces forty enquiries and twelve meetings. Initial volume favours the first, but the second deserves attention for its commercial progression. Cost, fit, proposals and sales still need reviewing before deciding a budget.
For long processes, group requests by entry date and follow their development. Do not compare all this month’s sales with all this month’s new contacts as if they necessarily involved the same people.
At the weekly review, choose one finding and one action: correct a link, improve the sales response or clarify the offer. Record the hypothesis and change date. At the next review, check the effect using comparable data and retain uncertainty when volume is insufficient for a conclusion.