E-commerce: how to allocate your budget between Meta Ads and Google Ads based on margins and new customers?

When we take over an e-commerce account, the split between Meta and Google often dates back to the store's launch. They started on Meta, added Google a few months later, and the percentages haven't changed since. No one really knows if the next euro would be better spent on one or the other.

To decide, you need two pieces of information that the platforms don't give you: your actual margin per order and the cost of a customer who has never bought from you before. This is the foundation upon which we build media plans for ourAds agency, for both Google and Meta. Here is the method, including the calculations and grids we use at Growth Room.

Key takeaways

• We allocate the budget based on the contribution margin, not the ROAS displayed in the platforms.
• The break-even ROAS is calculated by dividing 1 by the contribution margin rate.
• Google captures existing demand, while Meta creates new demand: they don't acquire the same customers.
• Our starting point: 40% to 50% on Google, 30% to 45% on Meta, 10% to 20% on recovery.
• At Poiscaille, this phase-based allocation generated over 2,000 subscriptions in just over two months.

Why shouldn't you rely on platform ROAS to allocate your budget?

The ROAS from Google Ads or Meta Ads Manager presents three problems.

• It doesn't account for your margin. A ROAS of 4 is excellent with a 60% margin, but only breaks even at 25%.
• It mixes existing and new customers. Campaigns targeting your brand name and retargeting show high ROAS because they reach people who already know you.
• Each platform claims the sale. If you add up the conversions reported by Meta and Google, you will often exceed your actual sales.

At EVA, a network of virtual reality centers we support, Google showed a ROAS of 9 and Facebook a ROAS of 2, for an overall ROAS of 6. Cutting Facebook would have reduced Google searches, which Meta was fueling. We managed based on cost per booking, brought down to 29% of the average order value.

How much can you afford to pay to acquire a customer?

Three figures are enough to answer that.

Step 1: calculate your contribution margin

Thecontribution marginis what remains from an order after accounting for product costs, shipping, returns, and payment fees. It is the budget available to cover acquisition costs.

Step 2: Determine your break-even ROAS

Thebreak-even ROASis calculated as follows: 1 ÷ contribution margin rate. Below this threshold, every sale results in a loss.

Step 3: Set a target CAC for a new customer

TheCACceiling for a new customer is equal to the margin from their first order, plus that of subsequent orders if your repurchase rate is tracked. The target CAC must remain below this to maintain a safety margin. Thissafeguardvaries based on the average order value:

• order value under €60: CAC between 30% and 40% of the order value;
• order value between €60 and €150: CAC between 25% and 35% of the order value;
• order value over €150: CAC between 20% and 30% of the order value.

Hypothetical example for a store with an average order value of €80:

In this example, a campaign that acquiresnew customersfor less than €24 can receive more budget, both on Meta and Google. Beyond €40, it costs more than it earns, unless your customers make repeat purchases regularly.

What roles should you assign to Meta Ads and Google Ads?

The two platforms reach different people at different times.

• Google Adsresponds topurchase intent: the user is already looking for a product like yours via Search, Shopping, or Performance Max. Conversion is fast, but volume depends on existing demand, and campaigns targeting your brand name primarily reach customers who already know you.
• Meta Adsintroduces your brand to people who weren't looking for you. This is often where the majority of new customers are acquired, though costs are more variable and success is highly dependent on creative assets.

Test budgets, the creative production schedule, and our six-week plan are detailed in our article ongenerating profitable e-commerce customers.

How do you adjust the split based on your profit margins?

With healthy margins, you can fund more discovery on Meta: the acquisition cost is higher, but your customer base grows faster. With tight margins, we prioritize Google first, where the intent is already present. We start with this framework and then adjust it based on account data:

Two scenarios change this framework:

• If your customers make frequent repeat purchases or subscribe, your CAC ceiling can incorporate theLTVcalculated over 6 or 12 months. You can then push Meta further to drive acquisition.
• If your priority is to increase the share of new customers, exclude existing customers from acquisition campaigns, cap the budget for brand campaigns, and track CAC for new customers rather than overall ROAS.

How do you distinguish between new and existing customers?

Without this distinction, Google will always appear more profitable than Meta. Here is what we implement for every account:

• Separate brand and non-brand campaigns on Google, with a capped budget for brand terms.
• Excluding customer lists from Meta and Google prospecting campaigns, and using dedicated audiences for remarketing.
• CAC per new customer for each channel: channel spend divided by attributed new customers, cross-referenced with the back-office or CRM.
• Daily monitoring of theMER(total revenue ÷ total media spend). It eliminates double attribution between platforms: as long as it remains above your target ROAS, the overall setup is profitable.
• Regular incrementality tests: turning off a portion of retargeting or reducing brand campaigns for three days to see what actually disappears from sales.

Poiscaille: a different distribution at each stage of the funnel

Poiscaille buys directly from French fishermen and sells fish basket subscriptions. The goal was to acquire subscribers at a cost compatible with their business model. We worked on Google Ads and Meta, adjusting the distribution based on the stage:

Between late November 2022 and late January 2023:

• over 2,000 subscriptions generated;
• a cost per subscription on Google that dropped from €77 in November to €6 in January;
• a cost per subscription on Meta that dropped from €112 in December to €27 in January.

These results only count one subscription per customer, even though subscribers stay for an average of 8 to 10 months. The actual profitability is therefore significantly higher, which justified investing so much in discovery on Meta.Read the full Poiscaille case study

At Editions Dupuis, whom we have been supporting since 2023, the same acquisition-led management approach delivers 230 new customers per month with a ROAS of 3. We have compiled other results in our summarye-commerce: maintaining performance when ROAS fluctuates.

How to readjust the budget each week?

The initial framework is a working hypothesis. Our consultants evolve it using this routine:

1. Rank campaigns by new customer CAC against the target CAC.
2. Cut the bottom 10% to 20% of underperforming offer, creative, and audience combinations.
3. Increase the budget by 20 to 30% for campaigns that have remained below the target CAC for at least three days.
4. Monitor the overall MER: if it drops below the break-even ROAS, slow down discovery and strengthen intent.
5. Test the incrementality of retargeting and brand campaigns once a month.

After a few weeks, the budget shifts toward the channel that acquires customers at the lowest cost, and the Meta / Google split establishes itself.

‍