Google catches, Meta creates: why an online store needs both channels

TL;DR

Google Ads works with people who are already searching for your product. Meta Ads shows the product to people who don’t yet know they need it. In LeadPrice practice, stores that combine both channels show a 40-60% higher ROAS than a single-channel strategy. For example, in the Adaptis case (premium e-commerce) Meta delivered a ROAS of 634%, and in another cosmetics project a ROAS of 12.75 over a year at $900K+ in revenue. Google meanwhile captured warm demand at a CAC 2-3x lower. But 70% of stores make the same mistake: they try to copy the strategy from one channel to the other, or pick just one “so as not to spread ourselves thin.” This article is about building a funnel where Google catches ready demand, Meta creates new demand, and both channels work toward a single business metric.

Why “either/or” doesn’t work for e-commerce

A typical situation: a store launches Meta Ads, sees a ROAS of 3-4, is happy for a month, then CPM rises 40-60%, the audience burns out, ROAS drops to 1.5-2. The owner thinks “Meta doesn’t work” and goes to Google Shopping. CAC is lower there, but traffic volume is 3-5x smaller — because Google only shows to those already searching. After 2-3 months growth hits the ceiling, because search demand is limited.

The real cause of the failure: the channels worked separately, not as a single system. Google captured 100% of the ready demand, Meta tried to generate cold traffic without retargeting the Google audience. The result — two separate funnels, neither of which sees the full picture of the customer.

At LeadPrice we’ve seen this in 7 out of 10 e-commerce clients at intake: they launched channels sequentially, not in parallel. When we rebuilt them into a single funnel (more on that below), the average ROAS gain was 45-70% in the first 3 months.

Framework: how Google and Meta work together (5 steps)

Here’s our methodology for building a two-channel funnel for an online store. It isn’t “launch both channels and wait,” but a structured process with clear economics at every step.

Step 1: Segment by demand temperature

What we do: Divide the audience into 4 segments by purchase intent.

• Hot — searching for your product right now (Google Search, Shopping)
• Warm — visited the site, viewed a product, didn’t buy (Meta retargeting + Google Display)
• Cool — know the category but don’t know you (Meta lookalike, Google Performance Max with a seed audience)
• Cold — don’t know they need the product (Meta creatives built around the problem)

How it works: Google takes the hot and part of the warm. Meta takes the cool and cold, plus aggressive retargeting of the warm. The key: don’t mix up the messages. The hot segment doesn’t need the problem explained — it needs the price, delivery, reviews. The cold segment is the opposite: first the pain point, then the product.

What this gives you: CAC on hot traffic (Google Search) is in practice 2.5-4x lower than on cold traffic (Meta prospecting). But hot volume is limited. In an average store with $50K/mo in revenue, search demand delivers 30-40% of sales, Meta 50-60%, the rest is organic and repeat.

Step 2: A 60/40 budget split with a feedback loop

What we do: The starting split is 60% Meta (demand generation), 40% Google (demand capture). But it isn’t a constant.

How it works: When Meta warms up an audience, some people don’t buy right away — they go and google the brand or the category. If your ads aren’t there, you lose 30-40% of potential customers. So increasing the Meta budget automatically requires strengthening Google (especially branded campaigns).

What this gives you: In LeadPrice cases, stores that synchronized their budgets showed a 25-35% rise in conversion with no change in total ad spend. Simply because they stopped losing warm traffic between the channels.

Step 3: Unified attribution (data-driven, not last click)

What we do: Switch off last-click attribution. Switch on the data-driven model in Google Ads (if there are 3,000+ conversions in 30 days) or linear/time decay. In Meta — 7-day click + 1-day view. The main thing is to link both channels via UTM + server-side tracking (Conversions API for Meta, enhanced conversions for Google).

How it works: The customer sees a creative on Meta → no click → 2 days later googles the brand → buys via Google Search. In a last-click model Meta gets 0 credit, Google 100%. That’s a lie. A data-driven model distributes the value: Meta 40%, Google 60% (say). Now you see each channel’s real contribution.

What this gives you: You stop underfunding Meta (which generates demand) and overfunding Google (which only collects it). In practice, after implementing correct attribution, Meta campaigns get +20-30% budget, ROAS drops from 5 to 3.8, but total profit grows 40% because you finally see the full picture.

Step 4: Lookalike audiences from Google Ads in Meta

What we do: Export the list of buyers from Google Ads (or the CRM, if there’s an integration) → upload it to Meta as a seed audience → build a 1-3% Lookalike.

How it works: People who buy via Google Search are an already validated audience. They have money, interest in the category, trust in online shopping. Meta takes their characteristics (demographics, behavior, interests) and finds similar people. The result is a cold audience, but of 40-60% higher quality than broad targeting.

What this gives you: A lookalike of Google buyers in our cases shows a CPA 30-50% lower than a lookalike of Meta buyers. The reason: the Google audience has higher lifetime intent — they were actively searching, not stumbling across an ad.

Step 5: Cross-channel retargeting (the hard kind)

What we do: Build a 4-tier retargeting ladder:

1. Google Display Remarketing — show all site visitors (from any source) static banners of the product they viewed + a 5-10% discount
2. Meta Dynamic Ads — show the catalog of products they viewed + similar ones (works better than static in 60% of niches)
3. Google Search Remarketing (RLSA) — raise bids on branded + category queries for those who’ve already been on the site (conversion is 2-3x higher)
4. Meta Lead Ads for abandoned carts — if someone added a product to the cart but didn’t check out, show a 3-field form right inside Facebook (name, phone, comment) → a manager calls

How it works: The person sees your product 4-7 times in different formats over 7-14 days. It isn’t spam, it’s a reminder. In e-commerce the average decision cycle is 3-10 days. If you disappeared after the first click, the customer went to a competitor.

What this gives you: Retargeting delivers 25-40% of all sales on 10-15% of the budget. Retargeting ROAS is 8-15 (3-4x higher than prospecting). In the Adaptis case with its 634% ROAS it was precisely retargeting that pulled the metric up — without it the overall ROAS would have been ~4.

Example in practice: a sports nutrition store

The client: an online sports nutrition store, revenue $80K/mo, average order $45, margin 40%. Before us they worked only with Google Shopping (CAC $18, ROAS 3.2). They wanted to scale, but search traffic had hit the ceiling.

What we did (the first 90 days):

• Launched Meta Ads with a $2,500/mo budget (60% of the Google budget)
• Segmented the audience: hot (Google Shopping + branded Search) → warm (retargeting in both channels) → cold (Meta prospecting on athletes’ pain points)
• Created 12 creatives for Meta: 6 on the problem (fatigue after training, slow muscle growth), 6 on the product (composition, certificates, reviews)
• Implemented the Conversions API for Meta + enhanced conversions for Google
• Created a 2% lookalike of Google buyers → it became the #1 audience in Meta with a CAC of $22 (versus $34 on broad targeting)

The result after 6 months:

• Revenue grew from $80K to $135K/mo (+69%)
• Google: CAC $18 → $16 (thanks to higher-quality branded traffic from Meta), ROAS 3.2 → 4.1
• Meta: CAC $28 (blended), ROAS 2.8 (on prospecting) + 11.3 (on retargeting)
• Overall blended ROAS: 5.4
• Share of repeat purchases: 18% → 31% (Meta retargeting + email sequences)

The key insight: without Meta the store would have been stuck at $85-90K/mo. Google had exhausted the search demand. Meta created new demand, part of which flowed into Google (branded queries +140% in 6 months), and the rest bought directly from Meta. The two channels reinforced each other.

When the framework isn’t the right fit (honestly)

This strategy is NOT for everyone. Here are 4 situations where it’s better to focus on one channel:

• Revenue under $20K/mo — not enough data to optimize two channels. Start with Google (if there’s search demand) or Meta (if the product is new), get to $30-40K, then add the second channel.
• Margin under 30% — CAC on Meta is usually higher than on Google. If your margin is 15-20%, Meta prospecting will eat all the profit. Work only with Google + organic until you’ve optimized the unit economics.
• An ultra-niche B2B product — if your target audience is 500 people in Ukraine, Meta won’t be able to find them (the algorithm needs a minimum of 5,000-10,000 to learn). Google Search + LinkedIn Ads is your path.
• No resources for creatives — Meta needs 8-12 new creatives every 30 days (because of fast burnout). If you have no designer or budget for shoots, Google Shopping + Performance Max will be more effective (product photos from the site work there).

At LeadPrice we tell clients honestly when it’s too early to launch a second channel. Better to run one channel at 8/10 than two at 4/10.

What you should do: a 14-day checklist

If only Google is running right now:

1. Export the list of buyers from the last 90 days from Google Ads or your CRM (at least 500 people)
2. Upload it to Meta as a custom audience
3. Create a 1-3% lookalike
4. Launch a test campaign on this lookalike with a $30-50/day budget, 4 creatives (2 on the problem, 2 on the product)
5. After 14 days look at the CAC — if it’s 2x lower than LTV, scale to 50-60% of the Google budget

If only Meta is running right now:

1. Do keyword research in Google Keyword Planner or Ahrefs — how many people search for your category + brand each month
2. If the volume is over 1,000 queries/month, launch Google Shopping + branded Search
3. Budget: 30-40% of Meta (no more at the start)
4. After 30 days look at how many branded queries came in — if it’s +50% on organic, Meta is creating demand and Google is collecting it. That’s the signal to increase both channels.

If both are already running, but separately:

1. Implement the Conversions API (Meta) + enhanced conversions (Google) — that’s 70% of the attribution work
2. Create cross-channel audiences: Google Display shows to those who clicked on Meta (and vice versa)
3. Synchronize discounts and promotions — if a 15% discount is running on Meta, it must also be in Google Shopping (otherwise the customer sees a discrepancy and loses trust)
4. Look at blended ROAS (not per channel) — if it’s above 3.5-4, everything is working

FAQ: Google Ads + Meta Ads for e-commerce

What’s the minimum budget to launch both channels?

A minimum of $1,500-2,000/mo in total ($900-1,200 Meta + $600-800 Google). Any less and there isn’t enough data to optimize the algorithms. Meta needs at least 50 conversions a week to work stably, Google Shopping 30-40. On a $500/mo budget you’ll get 10-15 conversions, the algorithm won’t learn, and CPM will be inflated. Better to start with one channel at $1,000, bring it to a stable ROAS of 3+, then add the second.

How do you know it’s time to add the second channel?

3 signals: (1) the current channel’s ROAS has been stable for 2-3 consecutive months (fluctuations under 15%), (2) you’ve exhausted the scaling options — raising the budget 30% brings only a 10-15% rise in sales, (3) CAC has started to rise at the same traffic volume (a sign of audience burnout). At that moment the second channel delivers maximum effect — it opens up a new audience or captures the demand the first one is creating.

Can cross-channel attribution be automated?

Yes, but not out of the box. You need either Google Analytics 4 with configured events + data-driven attribution (works with 3,000+ conversions in 30 days), or a separate platform like Supermetrics or Windsor.ai at $200-500/mo. The alternative is server-side tracking via a GTM Server Container + BigQuery, but that already requires a developer. At LeadPrice we set this up as part of the end-to-end analytics package — the client gets a single dashboard showing each channel’s real contribution to final profit, not just last-click conversions.

What mistakes are most common when launching two channels?

The top 3: (1) Copying the strategy — taking creatives from Meta and dumping them into Google Display. Doesn’t work. Google Display needs different formats, messages, CTAs. (2) No retargeting between channels — running prospecting in both but not showing the Meta audience ads in Google and vice versa. That loses 30-40% of conversions. (3) Different promotions in different channels — a 20% discount running on Meta, the old price in Google. The customer sees the discrepancy, thinks they’re being deceived, goes to a competitor. Synchronize everything: prices, discounts, messages, landing pages.

How long until a cross-channel strategy shows results?

The first 30 days are algorithm learning; ROAS will be 30-50% below target. Months 2-3 — stabilization, ROAS reaches a plateau. Month 4+ — scaling and optimization. You’ll see the real rise in business metrics (revenue, profit) on day 60-90. If blended ROAS is below 2.5 on day 90, something’s wrong with the funnel, product, or targeting. In our cases 80% of projects reached the target ROAS of 4-6 by the end of month 4. The other 20% were either a highly competitive niche (beauty, clothing) or unit-economics problems that advertising can’t fix.

Google Ads or Meta Ads — which to launch first for a new store?

It depends on whether there’s search demand. Go into Google Keyword Planner, enter 10-15 key queries for your niche (the category name, top products, branded if you have any). If the total volume is over 2,000-3,000 queries/month, start with Google Shopping — results come faster and CAC is lower. If there’s no demand (a new product, a narrow niche) — Meta, but be prepared to test 8-12 creatives and wait 60-90 days for payback. Ideally, if you have a $2,000+ budget, launch both right away, even for a new store. Google will collect whatever demand exists, Meta will start creating it. After 90 days you’ll get the synergy.

Summary: a two-channel funnel isn’t a luxury, it’s a necessity

If you want $100K+/mo in revenue, one channel isn’t enough. Google catches ready demand — but that demand is limited. Meta creates new demand — but without capturing it in Google you lose 30-40% of potential customers who go and google after seeing the ad. Both channels together produce synergy: a 40-70% rise in ROAS, a 20-30% drop in CAC, access to new audience segments.

The key: these aren’t two separate channels, it’s a single funnel with shared attribution, cross-channel retargeting, and synchronized messages. At LeadPrice we build such a funnel from scratch in 60-90 days — from a unit-economics audit to launching both channels with a single dashboard. If your store does $30K+/mo and you’re ready to scale, this is your next step.

Want to break down your situation? Fill in the brief — we’ll analyze your current channels, show where you’re losing money, and give you a 90-day roadmap. The first consultation is free, because we only take clients we’re sure we can help.

More cases from e-commerce and other niches — on the page of our projects. There we break down the numbers, mistakes, and solutions without embellishment.