Introduction
Return on ad spend is the metric most brands look at first. It is clean, comparable, and easy to communicate — a single number that appears to summarise whether your marketing is working.
The problem is that ROAS is a rear-view mirror. It tells you what happened inside your ad account during a specific period. It does not tell you why leads are not converting downstream. It does not reveal that your SEO traffic has strong intent but a poor landing page experience. It does not show you that 60% of your revenue comes from 15% of your customers, and that your retention spend is almost zero. It cannot flag that your attribution model is giving Google Ads credit for conversions that were actually driven by organic search.
Marketing analytics, done properly, is not about monitoring a single metric — it is about building a connected view of your growth system that tells you where to invest next, not just how last month performed.
This article covers the metrics and frameworks that growing brands — particularly those scaling past the point where intuition alone is enough — should be tracking, and why each one matters.
The Limits of ROAS as a Primary Metric
ROAS measures revenue generated per unit of ad spend. At its simplest: if you spend RM10,000 on ads and generate RM40,000 in attributed revenue, your ROAS is 4.0.
The number looks clear. In practice, it contains several layers of ambiguity that make it a poor standalone decision-making tool.
Attribution is rarely clean. Most customers touch multiple channels before converting — they may see a TikTok ad, later search for your brand on Google, then convert through a Meta retargeting campaign. Which channel gets the credit? Last-click attribution (the default for many platforms) gives all credit to the final touchpoint, which systematically undervalues upper-funnel channels like content and SEO. ROAS built on last-click data is not wrong — it is just incomplete.
Platform ROAS and business ROAS are different. Google Ads will report a ROAS based on the conversion value tracked through its pixel. That number does not account for returns, cancelled orders, low-margin products, or the cost of serving those customers. A campaign with a 5.0 ROAS on Google may be producing significantly lower actual margin contribution once those factors are included.
ROAS does not capture lifetime value. A customer acquired at what appears to be a poor ROAS may become one of your most valuable over twelve months. A campaign that looks efficient by ROAS may be acquiring one-time buyers who never return. Without LTV data, you cannot tell the difference.
The Metrics That Fill the Gaps
Cost Per Acquisition (CPA) and Cost Per Lead (CPL)
Where ROAS measures revenue output, CPA and CPL measure the cost of producing a specific outcome — a lead, a trial sign-up, a first purchase. These are particularly important for businesses with longer sales cycles or considered purchase decisions, where revenue attribution is delayed or indirect.
Tracking CPA and CPL over time — and breaking them down by channel, campaign, audience, and creative — gives you the granularity to allocate budget intelligently. A blended CPL number is almost always misleading: within it, some channels are performing at half the average and others at twice. Marketing analytics surfaces that variance so you can act on it.
Conversion Rate by Stage
A conversion rate is not a single number — it is a chain. Traffic to landing page. Landing page to lead. Lead to qualified conversation. Qualified conversation to closed customer. Each stage has its own conversion rate, and a drop at any stage has a different cause and a different fix.
Monitoring conversion rates across the full funnel — not just the top-of-funnel click-through rate — identifies precisely where the system is leaking. A landing page converting at 3% when the benchmark for your category is 7% is a more actionable insight than knowing your ad CTR is above average.
Customer Lifetime Value (LTV)
LTV is the total revenue a customer is expected to generate across their relationship with your business. It is the metric that determines how much you can rationally spend to acquire a customer — and therefore the ceiling on your acceptable CPA.
Brands that do not track LTV tend to under-invest in acquisition because their short-term ROAS or CPA benchmarks do not account for repeat purchase behaviour. They also tend to under-invest in retention, because they have no visibility of how much revenue they are leaving on the table by not nurturing existing customers.
LTV:CAC Ratio
The ratio of customer lifetime value to customer acquisition cost is one of the clearest indicators of whether a business is growing sustainably or burning through budget to acquire customers it cannot profitably serve. A ratio above 3:1 is generally considered healthy. Below 1:1, you are spending more to acquire customers than they will ever return.
This metric is particularly useful for scaling decisions. If your LTV:CAC ratio is healthy and improving, aggressive acquisition spend makes sense. If it is declining, the problem is almost certainly upstream — in lead quality, conversion rate, retention, or all three.
Channel Contribution and Assisted Conversions
Most marketing analytics setups overweight the last touchpoint and underweight everything that preceded it. Assisted conversion data — available through GA4 and multi-touch attribution models — shows which channels are contributing to conversion journeys even when they are not the final click.
SEO and content marketing are consistently undervalued in last-click models because they tend to operate earlier in the buyer journey. Paid social is similarly undervalued when it functions as a discovery channel that later converts through branded search. Without assisted conversion data, budget allocation decisions systematically defund the channels doing the most important work.
Return Rate and Margin by Channel
Not all revenue is equal. A campaign driving high-volume, low-margin purchases or a high rate of returns may appear strong on ROAS while contributing very little to actual profitability. Connecting your marketing analytics to product-level margin data and return rate data gives you a materially more accurate picture of which campaigns are actually generating business value.
Building a Marketing Analytics Infrastructure That Works
Tracking the right metrics requires the right infrastructure. A few principles that matter:
Server-side tracking is no longer optional. Browser privacy restrictions, iOS updates, and ad blockers have significantly reduced the reliability of client-side pixel data. Server-side solutions — including Facebook CAPI and server-side Google Tag Manager — capture conversion signals more accurately and feed better data into platform bidding algorithms. Clean data is the foundation of good analytics.
A single source of truth for cross-channel data. When each platform reports its own performance independently, the numbers never reconcile — Google, Meta, and TikTok will each claim credit for more conversions than actually occurred. A unified analytics setup — pulling data from all channels into a single dashboard — is what makes cross-channel comparison and attribution meaningful.
Segmentation by audience, channel, and product. Blended numbers hide variance. The most useful marketing analytics are segmented: CPL by channel and audience, conversion rate by traffic source and device, LTV by acquisition channel and product category. The more granular your view, the more precise your interventions.
Trend analysis over snapshots. A single month's data rarely tells you anything actionable. Marketing analytics becomes genuinely useful when you can see performance trends over time — rising CPL over six months, declining landing page conversion rate following a site update, improving LTV following a CRM automation implementation. Trend visibility is what allows you to intervene before a small problem becomes a large one.
What This Looks Like With Toggle
Toggle's reporting and business intelligence work with clients is built around connecting channel performance to commercial outcomes — not delivering platform dashboards that report in isolation.
For CIMB Malaysia, implementing a content-led SEO strategy required tracking not just keyword rankings, but organic impressions, organic click growth, and topical authority development over a 12-month period. The outcome — 52.6% increase in organic impressions and 34.5% growth in organic clicks — was visible because the right metrics were being tracked from the start of the engagement.
For UNITAR, tracking CPL across platforms simultaneously — with Facebook CAPI in place for accurate Meta attribution — made it possible to identify where spend was most efficient and scale accordingly. The 47% year-on-year CPL reduction was not a single campaign win; it was a tracking and optimisation system producing compounding results.