The Google vs Meta split, the 15-20% scaling rule, and the three checks you need to pass before you increase spend on a single campaign.
By Jack Goldsmith, Founder & Performance Marketer, Social Surge · 28 July 2026
What is the recommended budget allocation for scaling digital storefront ads?
Run a minimum of £1,000-£3,000 per month on your primary channel (Google Shopping first) until you reach 30-50 conversions per month at or below your target CPA. Then scale in 15-20% increments every 10-14 days. Add Meta once Google is consistently profitable, starting at roughly 30% of your Google spend and focused on retargeting before prospecting. Set your ROAS floor from your own margin data before you touch any budget lever.
The recommended budget allocation for scaling digital storefront ads is not a fixed percentage split you can lift from a benchmark table. It depends on your conversion volume, your margin, how mature your tracking is and which channel is already performing. The framework below is what we apply in the accounts we manage: the decision criteria, the channel sequencing and the scaling pacing that keeps performance stable as spend increases.
Before adjusting any budget, you need three green lights. In the accounts we manage, we will not recommend a scale-up unless all three are in place.
1. Tracking is clean. If GA4 and your Google Ads conversion tag are both firing on the same purchase event, your conversion volume is overstated and your reported CPA is fictionally low. Scaling on bad data just amplifies the illusion. Run a cross-check: compare GA4 transactions against your back-end order management system for the past 30 days. Any variance above 10-15% needs resolving before budget moves. Google's own conversion tracking guidance is explicit that duplicate tags are a common source of inflated conversion counts.
2. You have sufficient conversion volume. Smart Bidding strategies (Target ROAS, Target CPA, Maximise Conversion Value) require data to learn. Google's Smart Bidding documentation recommends a minimum of 30-50 conversions per month at the campaign level for tROAS to operate reliably. Below that threshold, the algorithm is pattern-matching on too small a sample and producing noisy, inconsistent results. Scaling a data-starved algorithm with more budget does not accelerate learning; it accelerates waste.
3. You are hitting your target at current spend. If you are spending £800 per month and your ROAS is 3x when your margin requires 6x, adding budget will not fix the ROAS gap. The gap is structural: wrong bids, wrong campaign structure, wrong product feed, or wrong audience signals. Scaling profitable campaigns is sensible. Scaling unprofitable ones just buys you more evidence that something is broken.
For e-commerce, Google Shopping and Performance Max capture demand that already exists. Someone searching "carbon road bike 54cm" or "barbel rig tackle set" has purchase intent you cannot manufacture. Meta, by contrast, creates demand: it introduces your product to people who were not actively looking for it.
In the accounts we manage, Google typically absorbs 55-70% of total paid media budget, with Meta at 30-45%, depending on product type, margin and how developed the Meta creative library is. Specialist niches (cycling, fishing, outdoor hobbyist gear) skew even heavier toward Google Shopping because the search intent is specific and high-value, and the audience volumes on Google are large enough to drive meaningful scale without needing Meta to fill gaps.
The practical implication: if you are starting from scratch, Google Shopping should be live and consistently profitable before you commit meaningful spend to Meta. A common mistake is splitting £2,000 per month 50/50 between both channels before either has enough conversion data to learn, producing two underperforming campaigns instead of one working one. Start narrow and deep, not broad and thin.
Once Google is generating 30-50 conversions per month at your target CPA, Meta Ads becomes the logical next channel. The initial Meta allocation should be modest, typically 20-30% of whatever Google is currently spending, and focused on retargeting (warm audiences, past visitors, add-to-cart abandoners) before prospecting into cold audiences.
A working split for a store spending £3,000 per month total might look like this:
The rationale for prioritising Meta retargeting is attribution overlap. A customer often clicks a Google Shopping ad, leaves without buying, then converts three days later after seeing an Instagram retargeting ad. Google claims the conversion in its last-click window; Meta claims it via view-through attribution. Neither is the complete picture. Running both channels and measuring blended ROAS across your total ad spend gives a more honest read of what your budget is actually producing. Our case studies page shows what this two-channel model looks like in practice across different niches.
Performance Max and Smart Bidding campaigns are sensitive to abrupt budget changes. When you double a campaign budget overnight, the algorithm interprets the spike as a significantly changed environment and enters a learning phase that temporarily degrades performance. In the accounts we manage, we typically see ROAS drop for 7-14 days after a large budget increase before it recovers to its pre-change level or better.
The safer method: increase budgets by no more than 15-20% every 10-14 days, then monitor CPA and ROAS for at least 7 days after each change before deciding on the next increment. For cycling and outdoor brands we manage, this pacing keeps performance stable through scale and avoids the pattern where a well-performing campaign falls apart because someone doubled the budget on a Friday afternoon.
For predictable seasonal peaks (Christmas, spring cycling season, bank holiday fishing weekends), plan the budget ramp-up 4-6 weeks in advance. A campaign that has been incrementally increased to £5,000 per month through November handles December peak budget far better than one jumping from £1,500 to £5,000 on 1 December. The algorithm has context; it does not cope well with shocks.
Every scaling decision should be anchored to a target ROAS built from your actual margin, not from industry averages or what a competitor claims to achieve. The simple version: if your blended product margin is 40% and you need a 10% net return on ad spend after the agency fee, your minimum viable ROAS is approximately 2.5x. Below that, every pound of ad spend is buying revenue at a loss.
In the niche verticals we operate in, our accounts typically run at 13x or above on Google for cycling clients, 12x or above for fishing, and 14x or above for outdoor gear. Those figures reflect a mature campaign structure with sufficient conversion history. They are not the starting expectation for a new account. When we onboard a new client, we set a conservative initial tROAS slightly below the required margin floor to allow the algorithm to gather clean data over the first 60-90 days, then tighten it progressively as volume and stability grow. You can see our published service fees on the pricing page to understand what the total cost-of-advertising picture looks like before setting your floor.
The rule: set your ROAS floor, then only scale budget when you are comfortably above it, not when you are touching it. A 10-15% buffer above your floor gives the algorithm headroom to optimise without immediately breaching profitability if performance dips during a scale-up.
Knowing when to hold is as important as knowing when to push. Do not increase ad budget when:
In our experience, premature scaling is the single biggest cause of well-structured campaigns losing performance. The conditions that led to the decision to scale are almost always visible in the data before the scale-up happens. If you want an independent view of whether your account is ready, a free PPC audit will flag the specific issues worth resolving before you increase any budget.
Most e-commerce stores need a minimum of £1,000-£1,500 per month on Google to generate 30-50 conversions per month in a reasonably competitive niche. Below that, campaign data is too thin for Smart Bidding to perform reliably. The specific number depends on your average order value, niche competition and product range.
A common starting split is 60-70% Google and 30-40% Meta, with Google leading until it is consistently profitable. Meta retargeting should be funded before Meta prospecting.
Increase budgets by no more than 15-20% every 10-14 days on Smart Bidding campaigns. Larger jumps trigger a learning phase and can cause ROAS to drop for 7-14 days before it recovers.
Set your minimum ROAS from your own margin data, not from industry averages. A 40% margin business typically needs at least 2.5x to break even on ad spend before overheads. Only scale comfortably above your floor, not when you are just touching it.
Book a free PPC audit: full account review, wasted-spend analysis and a clear growth plan, delivered in plain English.
Book Your Free Audit