Quick answer
Performance marketing KPIs that actually matter
The performance marketing KPIs that move revenue are ROAS net of returns, CAC, CPL, MQL→SQL conversion rate, and lead-to-close cycle time. Vanity metrics (CTR, CPM, reach) are useful diagnostics, not goals.
Pick one north-star KPI per account
Most accounts try to optimize 6 KPIs at once and end up optimizing none. We pick one — usually CAC for paid-only programs, CPL for lead-gen, ROAS net of returns for ecommerce — and align every weekly decision to it.
Supporting KPIs we always report
Cost per stage (visit → form → MQL → SQL → opp → closed), conversion rate per stage, time-to-conversion per stage, blended channel contribution, and CRM-graded lead quality.
Reporting cadence that compounds
Weekly performance dashboards for operators. Monthly business review for execs. Quarterly strategic review with revenue impact tied to specific decisions.
Glossary: the terms used in this guide
Performance marketing reporting breaks down when two people in the same meeting use the same word for different things. Before a dashboard is built, agree these definitions in writing and store them next to the dashboard. Most disputes between marketing, sales and finance in UAE companies trace back to an undefined term rather than a bad campaign. The list below is the minimum set. Add your own stage names if the sales process has more steps, but keep each definition to one sentence and tie it to a field in the CRM or analytics tool so it can be audited. The definitions should survive staff changes; a new marketing lead should be able to read them and produce the same numbers as the previous one. Where a term has a platform version and a business version, such as ROAS, name both explicitly so nobody mixes them. Revisit the list once a year, and whenever the sales process or the pricing model changes.
- CPL: media spend divided by leads received, where a lead is a confirmed form, call or chat with contact details.
- MQL: a lead that meets the written fit criteria, scored automatically or reviewed by marketing.
- SQL: an MQL that sales accepted and contacted, with a next step logged in the CRM.
- CAC: all acquisition spend, including media, agency fees and tooling, divided by new customers in the period.
- ROAS: revenue attributed to ads divided by media spend; net ROAS subtracts returns, cancellations and discounts first.
- LTV: gross margin expected from a customer over their lifetime, used to set the acceptable CAC ceiling.
- Payback period: months needed for a customer's gross margin to repay the CAC.
Worked example: turning a revenue target into weekly KPIs
Suppose a UAE software company wants AED 1.2 million in new annual contract value this quarter from paid channels, and the average contract is AED 40,000. That requires 30 new customers. If the CRM shows that one in five sales-qualified leads closes, the quarter needs 150 SQLs. If 40 percent of marketing-qualified leads become SQLs, it needs 375 MQLs, and if half of raw leads qualify, it needs 750 leads. Over 13 weeks that is about 58 leads, 29 MQLs and 12 SQLs per week. Those three weekly numbers become the operating KPIs. Budget follows from the chain: at an affordable CAC of AED 10,000, the quarter can spend AED 300,000, or roughly AED 23,000 a week, which implies a maximum CPL of AED 400. Each Monday the team checks the week against those targets. If leads are on target but SQLs are behind, the problem is qualification or sales follow-up, not media. If leads are behind and CPL is above AED 400, the problem is media. The value of the exercise is that every number has a named owner and a known tolerance before the quarter starts.
ROAS net of returns: why the platform number is not the business number
Ad platforms report ROAS as attributed revenue divided by spend, using their own attribution window and counting every order at checkout value. For UAE ecommerce that overstates the result in three ways. First, cash on delivery remains common, and COD orders are refused or returned at a materially higher rate than card orders, so a share of attributed revenue never arrives. Second, platforms count orders that would have happened anyway, particularly from branded search and from retargeting existing customers. Third, discounts and shipping costs are not subtracted. Net ROAS fixes this by taking accepted, delivered revenue after returns and discounts, then dividing by spend. The practical method is to pull delivered revenue from the order system monthly, map it to campaigns using order-level UTM or click IDs, and compare it with the platform figure. The gap is the correction factor; many stores find platform ROAS runs 30 to 60 percent above net ROAS. Apply the factor to weekly platform reporting so operators can still work in near real time, and recalculate it each month. A target ROAS set without this correction will look achieved while the business loses money.
Building the dashboard: GA4, platform data and the CRM in one view
A KPI dashboard that people trust is a plumbing project before it is a design project. Start by exporting GA4 to BigQuery so event-level data is available outside the interface and retained beyond the default window. Pull platform spend and clicks from Google Ads, Meta, LinkedIn and TikTok with a connector into the same warehouse, keyed by campaign name and date. Bring CRM records in with the stage timestamps, the original source and the click ID captured at form submission. Join the three on campaign name, which only works if a naming convention exists and is enforced; the convention is the single most important piece of the project. Build the dashboard in Looker Studio or an equivalent, with one page per audience: operators see cost per stage by campaign for the last seven and 28 days, executives see CAC, pipeline and net ROAS by month, finance sees spend against budget. Automate the refresh daily and write a short data dictionary that explains every column. Budget two to four weeks for the first version with a competent analyst, and expect to spend an hour a week maintaining it after that.
KPI mistakes we see in UAE accounts
Most KPI problems are not measurement errors; they are choices that made sense once and never got revisited. A target set when the account was new, a report built for a previous manager, an attribution setting changed during a platform update and never written down. The cost is slow: budget drifts toward campaigns that look good on the chosen metric and away from the ones that produce customers. Review the list below quarterly. If more than two apply, fix the reporting before adjusting any campaign, because changes made against a wrong KPI tend to make the account worse in ways that take months to surface. Each item has a simple test: ask whether a finance lead would accept the number as a basis for a budget decision. If the answer is no, the metric belongs in the diagnostics section, not in the headline, and the report should say so plainly.
- Reporting MQL volume as the headline while sales accepts fewer than a third of them.
- Comparing ROAS across channels that use different attribution windows and view-through settings.
- Using CTR or CPM as a goal, which rewards cheap, irrelevant clicks.
- Measuring CAC from media spend only, leaving out agency fees, tools and creative production.
- Changing the north-star KPI mid-quarter, which makes trend lines meaningless.
- Reporting blended CAC when branded search and existing customers are carrying the average.