Most eCommerce businesses start running paid ads before they have the foundations in place to make them work. The result is expensive traffic that does not convert and a growing conviction that paid media “does not work” for their store. This guide covers the right order of operations — foundation before scaling — and how to build a paid media strategy across Meta, Google Shopping, and TikTok that generates compounding returns rather than a recurring cost.
A scalable eCommerce paid media strategy requires three foundations before increasing ad spend: a store conversion rate above 2.5%, a live email programme that retains customers past the first order, and product pages with strong trust signals and page speed above 60 on mobile. Without those, every £1 in paid spend is working against itself. Once foundations are in place, allocate 60% to Google Shopping and Search, 30% to Meta, and 10% to TikTok or channel testing — then scale what the data shows working.
Suggested: Paid media channel mix diagram or Google Shopping / Meta campaign dashboard
Alt: “eCommerce paid media strategy channel mix showing Google Shopping, Meta ads and TikTok budget allocation”
The single most common paid media mistake in eCommerce is scaling ad spend before fixing the store. A Magento store converting at 0.8% on mobile will burn through paid budget at the same rate regardless of how well-targeted the ads are. The problem is not the ads. The problem is that 99.2% of visitors the ads send are leaving without buying.
Three foundations must be in place before increasing paid media spend produces a positive return:
A conversion rate below 2% on mobile means the store has structural UX, speed, or trust problems that paid spend cannot fix. Fix site search, product page hierarchy, checkout steps, and page speed first. See our guide to Magento conversion rate fixes for the full checklist.
Without post-purchase email flows, paid media pays to acquire customers once and loses them. With a live post-purchase sequence, welcome series, and win-back flow, the customer lifetime value from paid acquisition increases by 40 to 80%. See our guide to eCommerce email marketing for the seven flows that matter.
A product page without reviews, a clear returns policy, and delivery promise adjacent to the add-to-cart button will not convert paid traffic — even when that traffic has high purchase intent. Trust signals must be in place before paying to drive traffic to product pages.
The 5MS rule of thumb: for every £1,000 per month in paid media spend, the store should be generating at least £2,000 in customer lifetime value from retained customers — not just first-order revenue. If the maths does not work at your current retention rate, fix retention before increasing paid spend.
Google Shopping is the highest-intent paid channel available to eCommerce businesses. A buyer searching “men’s waterproof walking boots size 10” is at point of purchase decision. Google Shopping captures that moment with product listing ads showing image, price, and product name — directly competing on the search results page at maximum buyer intent.
| Criteria | Performance Max | Standard Shopping |
|---|---|---|
| Google control | High — Google chooses placements, audiences, creative | Low — you control bids, products, and search terms |
| Data requirement | Needs conversion history to optimise effectively (>50 conversions/month) | Works with limited data — better for new accounts |
| Best for | Scaling established accounts with strong conversion data | New accounts, controlled testing, brand protection |
| Transparency | Limited — search term visibility reduced | Full search term report, negative keyword control |
The quality of your Google Merchant Centre product feed determines how well Google Shopping performs, far more than bid strategy. A feed with rich product titles, accurate GTINs, detailed attributes, and up-to-date pricing and availability consistently outperforms a sparse feed with higher bids. Optimise the feed first:
- Product titles: include brand, key attributes, size, colour, and material in the title — not just the product name. “Berghaus Men’s Waterproof Walking Boots Gore-Tex Size 10 Black” outperforms “Hillmaster Boot” every time.
- GTINs and MPNs: products without GTINs or MPNs are deprioritised in Shopping auctions. Match every product to its manufacturer identifier.
- Custom labels: segment your catalogue by margin, best-seller status, and season using custom labels. Bid higher on high-margin products, lower on clearance.
- Negative keywords: in Standard Shopping, build a negative keyword list from the search term report every week. Blocking irrelevant queries is the fastest way to improve ROAS without touching bids.
Run branded Search campaigns alongside Shopping. Branded search has the highest conversion rate of any paid keyword type and protects your brand from competitors bidding on your name. It is also the cheapest traffic you will ever buy. Never leave branded Search to organic alone when running paid Shopping campaigns.
Meta ads (Facebook and Instagram) work differently from Google Shopping. Buyer intent on Meta is lower — people are not actively searching for your product. Meta is an interruption channel. Done well, it is a highly effective one. Done badly, it burns budget on audiences who will never buy.
Visitors who have viewed a product page, added to basket, or initiated checkout in the last 7 to 30 days. This is your highest-converting Meta audience. It is small but cheap and converts well. Run dynamic product ads (DPAs) showing the specific products each visitor viewed.
Lookalike audiences built from your customer purchase list, your email subscribers, or your top 10% value customers. Meta finds people who resemble your best buyers. Quality of the seed audience determines quality of the lookalike — use purchase data, not just website visitors.
Cold prospecting to interest-based or broad audiences. Highest spend, lowest conversion rate, longest payback period. Works for products with mass appeal and strong creative. Not appropriate as a starting point — run retargeting and lookalike first, then expand to cold prospecting once ROAS is proven.
On Meta, creative is the targeting. The algorithm finds buyers by showing your ad broadly and learning who clicks and converts. A strong creative — video or image that stops the scroll, communicates the product benefit in the first two seconds, and has a clear reason to act — will outperform precise audience targeting with weak creative every single time.
What works in 2026: user-generated content (UGC) style video, product demonstration videos, before/after content, and reviews as creative. Polished brand video with logos and music consistently underperforms authentic-looking content in Meta auctions. Test multiple creative formats. Kill what does not perform within two weeks. Scale what does.
TikTok ads for eCommerce represent a significant opportunity for the right products but are frequently misapplied. TikTok is not Google Shopping — buying intent is low and creative quality is everything. The stores that succeed on TikTok treat it as a content-led channel rather than a performance channel.
Products that show transformation, solve a visible problem, or have a satisfying result outperform on TikTok. Beauty, food, home, fitness, and novelty products consistently see strong TikTok performance. Complex B2B products or undifferentiated commodities do not.
TikTok’s core audience is 18 to 34. Products that skew outside this demographic — or that require significant purchase consideration — will not find an efficient audience on TikTok at scale.
TikTok Shop removes friction by enabling purchase without leaving the app. For consumer products, TikTok Shop consistently outperforms link-out ads to an eCommerce store. The integration requires a separate product feed and fulfilment workflow.
TikTok’s algorithm rewards content that looks like organic TikTok. Ads shot on phones in natural settings outperform studio-produced creative. Partner with micro-creators (10,000 to 100,000 followers) rather than spending on macro influencers for paid content.
Allocate no more than 10% of paid media budget to TikTok until you have proven ROAS from the channel. TikTok’s attribution is less reliable than Google’s, and the audience overlap with Meta means you can overcount conversions. Run incrementality tests before scaling TikTok spend significantly.
Budget allocation should reflect channel intent, not channel preference. Google Shopping captures existing demand. Meta creates demand. TikTok tests new audiences. Allocate proportionally to where buyers are in their journey.
| Channel | Allocation | Why | Optimise for |
|---|---|---|---|
| Google Shopping | 40% | Highest intent, captures buyers actively searching for your products | ROAS, conversion rate by product, search term relevance |
| Google Search (branded + key terms) | 20% | Branded protection and high-intent category queries | Impression share on branded, CPA on non-branded |
| Meta — retargeting | 15% | Warm audience, highest Meta conversion rate, dynamic product ads | Return on ad spend, frequency cap (max 3–4 per week) |
| Meta — prospecting | 15% | Lookalike audiences from customer list, new customer acquisition | New customer ROAS, cost per acquisition |
| TikTok / channel testing | 10% | New channel testing, top-of-funnel awareness, TikTok Shop | Incremental ROAS, view-through attribution with scepticism |
Rebalance quarterly, not daily. Paid media algorithms need time to learn. Making budget changes more frequently than every two weeks disrupts the learning phase and degrades performance. Set a 90-day review cycle for strategic budget reallocation and a weekly cycle for tactical bid adjustments.
Suggested: Paid media budget allocation pie chart or channel performance dashboard
Alt: “eCommerce paid media budget allocation framework showing Google Shopping 40%, Google Search 20%, Meta 30%, TikTok 10%”
Paid media, email marketing, and conversion rate optimisation are not separate workstreams. They are a system. Running them in isolation means each one underperforms relative to what it could achieve if connected.
Every paid landing page should have an email capture mechanism — exit-intent popup, scroll-triggered form, or basket offer. Paid traffic that converts to a subscriber but not a sale is not wasted if your email programme converts them within 30 days. Track this attribution in Klaviyo.
Upload your customer list to Meta and Google as a suppression audience. Stop paying to prospect people who already buy from you. Use that budget on lookalikes built from them instead.
Your Klaviyo high-value customer segment, your lapsed customer segment, and your VIP segment are all viable seed audiences for Meta lookalikes. The more specific the seed, the more valuable the lookalike. A lookalike built from your top 5% LTV customers consistently outperforms one built from all purchasers.
Paid traffic data is your most valuable source of CRO insight. The products with high click-through rate but low conversion rate from paid traffic have a product page problem — not a traffic problem. Use your paid media reports to identify which landing pages are losing traffic that paid to get there, then fix those pages first.
The most important paid media metric for eCommerce is not channel-level ROAS — it is blended ROAS across all paid channels combined. A campaign can show a strong Google ROAS while Meta is running at a loss, and the combined effect is neutral. Report blended ROAS every week and make budget decisions based on that number, not individual channel reports.
Attribution in eCommerce paid media is broken by default. Every platform over-claims credit. Google claims the sale. Meta claims the same sale. TikTok claims it too. The customer made one purchase.
GA4 last-click attribution is imperfect but consistent. Use it as the denominator for all channel performance comparisons. Never compare channels using their own platform attribution — they will all look great.
ROAS calculated on revenue is meaningless if your gross margin is 30%. Calculate ROAS on contribution margin — revenue minus cost of goods minus paid media spend. A 3× revenue ROAS at 30% margin is a 0.9× contribution ROAS — you are losing money.
Pause a channel for two to four weeks in a controlled test to measure true incremental revenue. Most eCommerce businesses discover that 20 to 40% of paid conversions would have happened organically. Factor that into your ROAS calculations.
Marketing Efficiency Ratio (total revenue divided by total ad spend across all paid channels) is the single metric that cuts through attribution noise. Track it weekly. It goes up as your store improves. It goes down when you overspend or when the store underperforms.
Scaling paid media spend before the conditions are right destroys margin. Scaling after the conditions are met compounds returns. The checklist before increasing paid budget significantly:
- Mobile conversion rate above 2.5% — if it is below this, additional traffic makes the problem worse, not better
- Post-purchase email flow live — customers acquired by paid spend need a retention programme or the acquisition cost is never recovered
- Blended ROAS above 2.5× at current spend level — if blended ROAS is below 2.5×, scaling spend reduces margin further
- Product feed fully optimised — Google Shopping with a rich, clean product feed consistently scales more efficiently than one with gaps
- Incrementality tested — you know what percentage of conversions are truly incremental, and your ROAS calculations account for it
When all five conditions are met, scale in 20 to 30% budget increments with a minimum two-week stabilisation period between each increase. Doubling budgets overnight disrupts campaign learning phases and typically causes a short-term ROAS decline even in healthy accounts.
The compounding advantage: eCommerce businesses that fix conversion rate, build email retention, and then scale paid media see compounding returns. Each improvement in CVR makes every £1 of paid spend more efficient. Each improvement in retention increases the LTV that justifies higher acquisition costs. The businesses that scale paid first and fix the rest later spend years paying to fix a system that was always working against them.
- Foundation before scaling. Fix conversion rate, email retention, and product page trust signals before increasing paid spend. Every structural improvement to the store makes paid media more efficient permanently.
- Google Shopping captures demand. Meta and TikTok create it. Allocate budget proportionally to buyer intent: 60% to Google, 30% to Meta, 10% to TikTok or testing.
- Product feed quality is the real Google Shopping lever. Rich titles, accurate GTINs, custom labels by margin, and weekly negative keyword reviews outperform bid strategy changes on a poor feed.
- Meta creative is the targeting. The algorithm finds buyers by learning from who responds. A strong UGC-style video with a clear benefit in the first two seconds outperforms precise audience targeting with weak creative.
- Measure blended ROAS and Marketing Efficiency Ratio, not channel-level ROAS. Every platform over-claims credit. GA4 last-click and total revenue divided by total ad spend are the metrics that cut through attribution noise.
- Paid media, email, and CRO are one system. Email suppressions improve paid targeting. Paid traffic data identifies CRO priorities. Email LTV justifies paid acquisition cost. Manage all three together.
5MS reviews eCommerce paid media strategies as part of a full-service growth audit — covering channel mix, budget allocation, feed quality, creative performance, and how paid media connects to your email programme and store conversion rate.
A scalable eCommerce paid media strategy requires three foundations — conversion rate above 2.5%, a live email retention programme, and optimised product pages — before increasing spend. Once in place, allocate 40% to Google Shopping, 20% to Google Search, 15% to Meta retargeting, 15% to Meta prospecting, and 10% to TikTok or channel testing. Connect paid media to email by suppressing existing customers from prospecting and using email segments as seed audiences for lookalikes. Measure blended ROAS and Marketing Efficiency Ratio — not channel-level metrics from platforms that all over-claim credit. Scale in 20 to 30% increments with two-week stabilisation periods between each increase.
Related guides from 5MS:
Common questions about eCommerce paid media strategy. Get in touch if yours is not here.
We review eCommerce paid media strategies every week. Bring us your current channel mix, your ROAS by channel, and your store conversion rate. We will tell you straight where the gaps are and what to fix first.