Performance Marketing Efficiency in Asia: How to Spend Less and Grow More

Performance Marketing Is Not a Growth Strategy. It Is a Distribution Channel.

Most scaling businesses in Asia treat performance marketing as their primary growth engine. They are wrong about the category. A distribution channel that stops working the moment you stop paying for it is not a growth engine. It is a rented audience, and the rent keeps going up.

This distinction matters enormously when you are allocating capital. A growth strategy builds compounding advantages: brand recall, organic discovery, community, pricing power. Performance marketing delivers volume on demand. Both are useful. Conflating them is expensive.

What follows is how we think about performance marketing efficiency at Elara Ventures, drawn from working with founders across Sri Lanka, India, and Southeast Asia who have had to make every marketing dollar work harder than their competitors.

---

What Marketing Mix Modelling Actually Tells You About Channel Efficiency

Marketing mix modelling (MMM) is the most underused analytical tool among growth-stage companies in South Asia. It attributes revenue contribution across every channel in your mix and, critically, it surfaces the diminishing returns curve for each one.

Most founders optimise on last-click attribution. That tells you which channel got credit for the final conversion. MMM tells you which channels actually moved a customer through their decision journey, and at what marginal cost. attribution modelling for Southeast Asian ecommerce

The diminishing returns curve is where the real insight lives. Every channel has a spend threshold beyond which each additional dollar buys you progressively less incremental revenue. Google Search keywords in competitive categories across India and Sri Lanka hit diminishing returns faster than most teams expect because branded and high-intent terms are finite. When you are bidding against the same five competitors for the same 10,000 monthly searches, your twelfth rupee is worth less than your first.

Building an MMM model does not require a data science team. It requires clean revenue data, channel spend data at a weekly granularity, and a willingness to run the regressions. A Colombo-based fashion retail brand we worked with built a workable MMM in a Google Sheet over eight weeks. The output told them their Meta spend had been operating past the point of diminishing returns for four months while their Google Shopping campaigns still had room to scale. They reallocated 30 percent of their Meta budget and recovered meaningful margin within a quarter.

---

CAC Payback Period by Cohort and by Channel: The Metric That Prevents Subsidised Losses

Average CAC is a dangerous metric. It flattens the performance of your best channels against your worst and produces a number that feels acceptable while hiding campaigns that are destroying value.

The metric that actually protects your capital is CAC payback period, tracked at the cohort level and at the channel level simultaneously. unit economics framework for scaling startups

CAC payback period answers the question: how many months of gross margin does it take to recover the cost of acquiring this customer? A customer acquired through an Instagram video ad in March and a customer acquired through a Google Shopping ad in March are two different cohorts, even if they bought in the same week. They will retain differently, repurchase at different rates, and ultimately produce different lifetime values.

The discipline is to define your acceptable payback window before you scale. For subscription businesses across Southeast Asia, we typically see founders target six to twelve months. For high-frequency consumer categories like beauty or grocery, the window should be tighter, closer to three to four months. For B2B SaaS businesses operating out of Colombo or Bangalore, eighteen to twenty-four months can be rational if your net revenue retention is above 110 percent. The window is a function of your business model, not a universal constant.

When you track CAC payback by channel, you will almost always find one or two channels that are structurally unprofitable by your defined window. The discipline is to cut them, not to average them away. A Sri Lankan logistics platform we advised was running seven paid channels simultaneously. Three of them had never recovered CAC within any reasonable window across eight months of data. Cutting those three and concentrating spend on the remaining four reduced blended CAC by 22 percent within ninety days.

---

How Nykaa and Mamaearth Built Performance Marketing Machines in India

The two most instructive performance marketing case studies in South Asian consumer commerce are Nykaa and Mamaearth. They took different paths but both demonstrate what rigorous attribution and spend discipline look like at scale.

Nykaa built its demand engine on Instagram and Google with ROAS tracking at the SKU and campaign level. This is unusual discipline. Most brands track ROAS at the campaign or ad set level. Nykaa went to the product level, which allowed them to identify which specific SKUs were profitable to advertise and which were not. A lipstick with strong organic search demand did not need paid support. A newly launched serum in a crowded category did. Spend followed that logic, not category-level assumptions.

Mamaearth's story is about competing asymmetrically. Established FMCG companies in India were spending hundreds of crores on television. Mamaearth had no access to that channel at comparable scale. So they built a digital performance marketing operation that could acquire customers at a cost structure the incumbents were not watching closely enough. They were not trying to win on share of voice. They were winning on share of consideration among digitally native consumers, and they were measuring every rupee of that spend against direct revenue outcomes.

Both companies eventually built brand equity on top of their performance foundations. That sequencing matters. Performance marketing funded the growth that created the brand. The brand then reduced their dependence on paid acquisition over time. That is the correct order of operations. brand building vs performance marketing for Asian startups

---

Blended ROAS: The Metric That Hides Unprofitable Campaigns

Blended ROAS is one of the most common analytical errors we see in growth-stage businesses across South and Southeast Asia. It averages your best campaigns with your worst and produces a headline number that looks healthy while subsidising real losses.

Here is how it works in practice. Your branded keyword campaigns on Google have a ROAS of 12x. They should. Customers searching for your brand name are already sold. You are essentially paying to intercept demand you largely created through other means. Your prospecting campaigns on Meta targeting cold audiences have a ROAS of 1.8x. At your gross margin, that is unprofitable. Blended, you are showing a 4x ROAS to your board and congratulating yourself on your marketing efficiency.

The fix is segmentation before reporting. Separate branded from non-branded search. Separate retention campaigns from acquisition campaigns. Separate warm retargeting from cold prospecting. Each segment needs its own ROAS floor, calibrated to the gross margin of the products being sold and the expected lifetime value of the customer type being acquired. marketing reporting frameworks for growth stage companies

We have seen this error in businesses across Colombo, Jakarta, and Chennai. It is not a market-specific failure. It is a reporting discipline failure. The data to fix it usually already exists. The problem is that no one has segmented it.

---

How Performance Marketing Dependency Destroys Full-Price Revenue

The second structural failure pattern is more insidious because it takes eighteen to twenty-four months to become visible. Running continuous promotional campaigns to acquire or reactivate customers trains your audience to wait for discounts. Once trained, they do not buy at full price.

This is a genuine threat to long-term unit economics. A Southeast Asian apparel brand may show excellent short-term ROAS on promotional campaigns. But if the cohort analysis shows that customers acquired through a discount event repurchase almost exclusively during subsequent discount events, those customers are structurally less valuable than customers who discovered the brand organically and convert at full price.

The measurement is straightforward: track the full-price purchase rate by acquisition channel and by acquisition offer type. If discount-acquired cohorts have a full-price purchase rate below 20 percent across six months, you are not acquiring customers. You are renting transaction volume and compressing your average selling price permanently.

The correction involves a deliberate reduction in promotional frequency, acceptance of a short-term volume decline, and a shift in spend toward content, community, and brand channels that attract customers with higher price tolerance. This is a hard conversation to have with a board that is watching monthly revenue. It is a necessary one.

---

Marginal CAC: The Number That Tells You When to Stop Scaling

Average CAC tells you what you have spent. Marginal CAC tells you what the next customer will cost. These numbers diverge materially as you scale, and ignoring that divergence is how companies grow themselves into unprofitability.

Marginal CAC rises because the most receptive customers convert first. Early in a campaign, you are reaching people with genuine purchase intent or existing brand awareness. As you exhaust that pool and move to broader audiences, conversion rates fall and cost per acquisition climbs. This is not a failure of execution. It is a structural feature of paid media markets.

The practical implication is that you cannot plan next quarter's marketing budget using this quarter's average CAC. You need to model where on the diminishing returns curve you will be operating at different spend levels and set your budget to the point where marginal CAC still sits below the threshold your payback period framework defines as acceptable.

A Bangalore-based D2C brand we worked with was planning to double their Meta spend based on a six-month average CAC of 480 rupees. When we modelled their marginal CAC at double the spend, it came out above 900 rupees. Their payback window made 480 rupees viable. It did not make 900 rupees viable. They did not double the Meta budget. They diversified into influencer partnerships and organic search instead.

---

FAQ: Performance Marketing Efficiency in Asian Markets

What is a good CAC payback period for a startup in South Asia?

The right payback period depends on your business model and gross margin. For consumer subscription businesses in India or Sri Lanka, six to twelve months is a reasonable target. For high-frequency D2C categories, aim for three to four months. For B2B SaaS with strong net revenue retention, eighteen to twenty-four months can be justified. The key is to define the window before you scale, not after.

How is marketing mix modelling different from last-click attribution?

Last-click attribution gives credit to the final touchpoint before a conversion. Marketing mix modelling estimates the revenue contribution of every channel in your mix using regression analysis on historical spend and revenue data. MMM also surfaces diminishing returns curves, showing you the spend level at which each additional dollar in a channel stops generating proportional revenue. It is more useful for budget allocation decisions.

Why does blended ROAS mislead growth-stage companies?

Blended ROAS averages your best-performing campaigns, typically branded search, with your worst, typically cold prospecting. The resulting headline number looks acceptable while masking campaigns that are destroying margin. To avoid this, segment ROAS by campaign type: branded versus non-branded, retention versus acquisition, warm retargeting versus cold audiences, and set floor thresholds for each segment based on your gross margin.

How do you stop performance marketing from training customers to only buy on promotion?

Track full-price purchase rates by acquisition channel and acquisition offer type. If discount-acquired cohorts show low full-price repurchase rates over six months, begin reducing promotional frequency and shifting spend toward brand and content channels that attract customers with higher price tolerance. Accept a short-term volume dip. The alternative is permanently compressed average selling prices across your customer base.

---

The Honest Summary: Performance Marketing Is a Tool, Not a Moat

Every performance marketing channel available to a startup in Colombo, Jakarta, or Chennai is available to their best-funded competitor. Spend efficiency is a real advantage but it is an executional advantage, not a structural one. The businesses that build durable growth in Asian markets use performance marketing to fund early customer acquisition while simultaneously building the brand equity, organic discovery, and community that reduce their dependence on paid channels over time.

That is the transition worth planning for. Performance marketing scales your customer base. Brand and product build the loyalty that makes that customer base worth having.