---
title: "Paid acquisition for e-commerce | Optimize All Academy"
description: "Paid media: fuel, not foundation Paid advertising can scale e-commerce quickly — but only when the foundations (product, pages, margins, retention) are…"
url: https://optimizeall.com/learn/ecommerce-marketing-and-growth/paid-acquisition-for-ecommerce
updated: 2026-10-05
---

E-commerce Marketing and Growth · Marketplaces, social commerce, product feeds and paid acquisition · lesson 14 of 20 · 12 min

# Paid acquisition for e-commerce

## Paid media: fuel, not foundation

Paid advertising can scale e-commerce quickly — but only when the foundations (product, pages, margins, retention) are sound. Otherwise, ads amplify losses.

## Core paid channels

| Channel | Role | Notes |
|---|---|---|
| Search and shopping ads | Capture existing demand | Product feeds, shopping campaigns, brand and generic search |
| Social ads | Create demand and retarget | Creative-driven; broad targeting with strong creative is common |
| Marketplace ads | Visibility within marketplaces | Measure against marketplace margin |
| Video and connected TV | Awareness at scale | Harder to attribute; consider incrementality tests |
| Affiliates and creators | Partner-driven sales | Covered in Module 5 |

## Product feeds

Shopping and catalogue ads rely on a **product feed**: a structured file of your products (title, description, price, availability, images, identifiers such as GTIN, categories). Feed quality strongly affects performance:

- Descriptive titles with key attributes.
- Accurate prices and availability (mismatches can cause disapprovals).
- High-quality images.
- Correct identifiers and categories.
- Custom labels (for example, margin tier, bestseller, seasonal) to structure campaigns.

## Creative is the new targeting

On social platforms, automated delivery systems increasingly find audiences based on signals from creative and conversions. As a result, **creative quality and variety** are major levers:

- Test multiple concepts (problem-solution, demonstration, testimonial, unboxing, comparison), not just colour variations.
- Use native formats (vertical video, captions, creator-style content).
- Refresh creative regularly to combat fatigue.
- Keep a creative log: concept, hook, format, results.

## Measurement: ROAS versus MER

```
ROAS (return on ad spend) = Revenue attributed by the platform / Ad spend (per channel/campaign)
MER (marketing efficiency ratio) = Total revenue / Total marketing spend (whole business)
```

Platforms attribute sales using their own rules and often overlap (the same sale claimed by several platforms). MER gives a blended view that cannot double count. Use **platform ROAS for in-platform optimisation** and **MER and contribution margin for business decisions**. For bigger budgets, incrementality tests (geo holdouts, conversion lift studies) show what ads truly add.

## Break-even ROAS

Know the minimum ROAS at which ads are profitable on the first order:

```
Break-even ROAS = 1 / Contribution margin % (before marketing)

Example: contribution margin before marketing = 40%
Break-even ROAS = 1 / 0.40 = 2.5
```

Below 2.5, the first order loses money (though repeat purchases may make it worthwhile — see Module 6). Set target ROAS based on your margin and customer lifetime value, not on generic benchmarks.

## Tracking foundations

- Implement conversion tracking properly (pixels/tags and server-side conversion APIs where appropriate), respecting consent requirements.
- Use UTMs consistently for non-auto-tagged channels.
- Reconcile platform-reported conversions with your store's orders.
- Track new versus returning customer revenue — acquisition campaigns should acquire new customers.

## Worked example: scaling a UK candle brand

Illustrative numbers: contribution margin before marketing is 45%, so break-even first-order ROAS is about 2.2. Social ads report a ROAS of 3.0, but MER across all spend is 2.4 and new-customer share is falling. The team:

1. Excludes existing customers from prospecting campaigns.
2. Tests new creative concepts (candle-making process, scent descriptions by creators).
3. Runs a geo holdout in two regions to measure incrementality.
4. Scales only campaigns that maintain MER and new-customer acquisition.

## Automated campaigns in 2026: what they need from you

| Campaign type | What it automates | Your inputs that matter most |
|---|---|---|
| Google Performance Max (with Merchant Center feed) | Bidding, placements and creative across Search, Shopping, YouTube, Display, Gmail, Maps, Discover | Feed quality, conversion values (ideally margin-aware), asset groups, audience signals, brand exclusions |
| Google Demand Gen | Visual ads on YouTube (incl. Shorts), Discover, Gmail | Creative variety, product feed for shoppable formats |
| Meta Advantage+ sales campaigns (renamed from Advantage+ shopping in 2025) | Audience, placement, creative combinations | Creative diversity, catalogue quality, Conversions API, clean purchase events |
| TikTok Shop ads (e.g. GMV Max, where TikTok Shop operates) | Budget across product and live content | Product listings, creator content, stock |
| Retail media (Amazon Ads, noon, Daraz, etc.) | Sponsored placements on the marketplace | Listing quality, reviews, price, stock |

## Hands-on: feed margin tiers into campaigns

Platforms optimise to the value you report. If every sale counts equally, they will happily sell your lowest-margin products. Two practical fixes:

```python
# Assign margin tiers to products for use as a custom label in your feed (illustrative thresholds)
def margin_tier(price: float, unit_cost: float, avg_fulfilment_cost: float) -> str:
    margin = (price - unit_cost - avg_fulfilment_cost) / price
    if margin >= 0.50: return "margin_high"
    if margin >= 0.30: return "margin_mid"
    return "margin_low"

print(margin_tier(price=189, unit_cost=60, avg_fulfilment_cost=18))   # margin_high
print(margin_tier(price=49, unit_cost=24, avg_fulfilment_cost=12))    # margin_low
```

1. Put the tier into a feed **custom label** (e.g. `custom_label_0`) and structure campaigns or asset groups so high-margin products get more aggressive targets.
2. Where your stack supports it, send **profit or margin-adjusted conversion values** rather than revenue (server-side via Google Ads enhanced conversions / conversion adjustments or Meta Conversions API) — keep consistent definitions and test carefully.

## Measuring what ads really add

- Use **MER** (total revenue ÷ total marketing spend) and new-customer revenue for business decisions; use platform ROAS for in-platform optimisation.
- Run **incrementality tests**: Meta Conversion Lift, Google Ads conversion lift where eligible, or geo holdouts (turn spend off in matched regions).
- Consider **marketing mix modelling** (MMM) once spend is large and spread across channels; open-source options include Google's Meridian and Meta's Robyn.

## Common mistakes

- Scaling spend on platform ROAS alone.
- Poor product feeds causing disapprovals and weak relevance.
- Retargeting-heavy budgets that claim credit for purchases that would happen anyway.
- Ignoring creative fatigue.
- No break-even calculation.

## Budget pacing and learning phases

Automated campaigns typically need enough conversions to learn. Frequent large budget changes or constant edits can reset learning and make performance unstable. Change budgets in measured steps, give campaigns time to stabilise, and group products or audiences so each campaign has enough data. During peak seasons, plan budget increases in advance rather than reacting hour by hour.

## Paid acquisition checklist

- [ ] Break-even ROAS calculated from contribution margin
- [ ] Product feed optimised with custom labels
- [ ] Creative testing programme with concept variety
- [ ] MER and new-customer revenue tracked weekly
- [ ] Incrementality tested before major budget increases

## Video lecture: Paid acquisition for e-commerce

Lecture coming soon · 15 chapters · about 9 minutes. Read the full transcript below.

1. Paid acquisition for ecommerce
2. Ads amplify what exists
3. Core paid channels
4. Automated campaigns
5. Creative is the new targeting
6. ROAS vs MER
7. Break-even ROAS
8. Retargeting: handle with care
9. Simple example: UK candles (illustrative)
10. Realistic example: Saudi electronics PMax (illustrative)
11. Watch me do it: margin tiers
12. Tracking + incrementality
13. Pacing and learning
14. Common mistakes
15. Recap

## Lecture transcript

### Paid acquisition for ecommerce

Paid ads can grow an online store faster than anything else. They can also lose money faster than anything else. The difference usually isn't the platform. It's whether your foundations, product feed, margins and measurement are ready. In this lecture you'll learn the core paid channels, what automated campaigns like Performance Max and Meta's Advantage+ sales campaigns need from you, why creative is the new targeting, the difference between ROAS and MER, break-even ROAS, and how to measure what ads really add. Then you'll watch me feed margin tiers into campaigns.

### Ads amplify what exists

Why be careful? Because ads amplify whatever you already have. A great product with good pages, healthy margins and customers who come back? Ads accelerate it. A weak product page, thin margins or poor retention? Ads accelerate the losses. And today's campaigns are largely automated, so the platform optimises towards whatever signal you give it. Feed it bad data, and it will efficiently find you the wrong customers.

### Core paid channels

The core channels. Search and shopping ads capture existing demand, powered by your product feed. Social ads create demand and retarget, driven by creative. Marketplace and retail media ads sell within marketplaces like Amazon, noon and Daraz. Video and connected TV build awareness at scale but are harder to attribute. And affiliates and creators, covered in the next module. Each has a different role, and a healthy mix captures demand and creates it.

### Automated campaigns

Automated campaigns now dominate. Google's Performance Max uses your Merchant Center feed to run across Search, Shopping, YouTube, Display, Gmail, Maps and Discover. Demand Gen runs visual ads on YouTube, including Shorts, Discover and Gmail. Meta's Advantage+ sales campaigns, renamed from Advantage+ shopping in twenty twenty-five, automate audiences, placements and creative combinations. Where TikTok Shop operates, its automated Shop ads allocate budget across product and live content. Your job has shifted. You no longer set every lever. You supply the inputs: a clean feed, accurate conversion values, varied creative, and sensible constraints.

### Creative is the new targeting

Creative is the new targeting. Automated delivery systems find audiences largely from signals in your creative and conversions. So creative quality and variety are major levers. Test genuinely different concepts: problem and solution, demonstration, testimonial, unboxing, comparison. Use native formats, like vertical video with captions and creator-style content. Refresh creative before fatigue sets in. And keep a creative log of concept, hook, format and results, so you learn which angles work for your customers.

### ROAS vs MER

Now measurement. ROAS, return on ad spend, is revenue attributed by a platform divided by its ad spend. MER, marketing efficiency ratio, is total revenue divided by total marketing spend across the whole business. Platforms attribute sales using their own rules, and several may claim the same order. MER can't double count. So use platform ROAS to optimise inside each platform, and MER, new-customer revenue and contribution margin to make business decisions.

### Break-even ROAS

Break-even ROAS tells you the minimum return at which the first order pays for itself. It's one divided by your contribution margin before marketing. With a forty percent contribution margin, break-even ROAS is two point five. Below that, the first order loses money, although repeat purchases may make it worthwhile, which is why lifetime value matters. Set targets from your own margins and retention, not from generic benchmarks.

### Retargeting: handle with care

Retargeting deserves special attention, because it often looks better than it is. Retargeting ads show products to people who've already visited your site, and they report impressive ROAS. But many of those people would have come back and bought anyway. The platform claims credit for a sale you'd have made. So cap retargeting budgets, exclude recent purchasers, keep frequency reasonable, and judge it with incrementality tests or holdouts. Put most of your growth budget into prospecting, which finds new customers. Retargeting should be a helpful reminder, not a machine for claiming credit.

### Simple example: UK candles (illustrative)

A simple example. A UK candle brand, with illustrative numbers, has a forty-five percent contribution margin before marketing, so break-even first-order ROAS is about two point two. Social ads report a ROAS of three. Looks great. But MER across all spend is two point four and falling, and the share of revenue from new customers is dropping. The team realises retargeting is claiming credit for returning customers. They cap retargeting spend, shift budget to prospecting with new creative, and judge success by MER and new-customer revenue.

### Realistic example: Saudi electronics PMax (illustrative)

Now a realistic scenario with illustrative details. A Saudi electronics retailer runs Performance Max with every sale counted equally. The campaign pushes cheap accessories with thin margins, because they convert easily. Revenue grows, but profit doesn't. The team assigns margin tiers to every product and adds them as a custom label in the feed. They split asset groups by margin tier, with more aggressive targets for high-margin products, and start testing margin-adjusted conversion values. They also exclude out-of-stock products automatically. The campaign's revenue grows more slowly, but contribution margin improves.

### Watch me do it: margin tiers

Watch me feed margin tiers into campaigns. I write a tiny Python function that takes price, unit cost and average fulfilment cost, calculates the margin, and returns high, mid or low tier. A hundred and eighty-nine dirham shirt costing sixty plus eighteen for fulfilment is high margin. A forty-nine dirham accessory costing twenty-four plus twelve is low margin. I run it across the catalogue, write the result into custom label zero in a supplemental feed, and structure campaigns or asset groups by tier. Where my tracking supports it, I test sending margin-adjusted conversion values server-side. Now the machine optimises towards profit, not just revenue.

### Tracking + incrementality

Tracking and incrementality. Implement conversion tracking properly, with pixels and server-side connections like Meta's Conversions API and Google's enhanced conversions, respecting consent. Reconcile platform-reported purchases with your store's orders every week. Track new versus returning customer revenue. And test what ads truly add with incrementality studies: Meta Conversion Lift, Google conversion lift where eligible, or geo holdouts where you turn spend off in matched regions. Once spend is large and spread across channels, consider marketing mix modelling; open-source tools include Google's Meridian and Meta's Robyn.

### Pacing and learning

Budget pacing and learning phases matter with automated campaigns. When you launch or make a big change, the system needs data to learn, so avoid changing budgets, targets and creative every day. Make changes in steps, for example raising budgets by modest amounts at a time, and give the campaign time to settle before judging. Consolidate rather than fragment: a few campaigns with enough conversions learn faster than many tiny ones. And plan for peaks, like White Friday or Ramadan, by increasing budgets ahead of time and checking stock, so the machine isn't learning in the middle of your busiest week.

### Common mistakes

Common mistakes. Scaling ads before fixing conversion or margins. Judging success by platform ROAS alone. Letting retargeting claim returning customers. Feeding automated campaigns poor product data. Counting all products as equally valuable. Refreshing creative too rarely. And never testing incrementality.

### Recap

Recap. Paid ads amplify what exists, so fix foundations first. Automated campaigns need a clean feed, accurate and ideally margin-aware conversion values, varied creative and sensible constraints. Use platform ROAS for optimisation, MER and new-customer revenue for decisions, and break-even ROAS to set targets. Prove what ads add with incrementality tests. Try this now: calculate your break-even ROAS, compare it with last month's MER, and run the margin tier function from the lesson text on your catalogue.

## Key takeaways

- Paid media amplifies whatever foundations you have — good or bad.
- Product feed quality and creative variety are major performance levers.
- Use platform ROAS for optimisation and MER plus contribution margin for business decisions.
- Calculate break-even ROAS as 1 ÷ contribution margin percentage.

## Try it

Calculate your store's break-even ROAS, compare it with current platform ROAS and MER, and list two creative concepts to test next month.

- [Previous: Google Merchant Center and product feeds](https://optimizeall.com/learn/ecommerce-marketing-and-growth/google-merchant-center-and-product-feeds)
- [Next: Designing an affiliate programme](https://optimizeall.com/learn/ecommerce-marketing-and-growth/affiliate-programmes)
- [All lessons of E-commerce Marketing and Growth](https://optimizeall.com/learn/ecommerce-marketing-and-growth)