Google Ads vs Meta Ads: 2026 Budget Split Guide

In Brief

Google Ads captures demand that already exists. Meta Ads creates demand that does not. For a Turkish ecommerce brand spending around 100,000 TL a month, a sensible 2026 starting point is 55 percent Google and 45 percent Meta. If nobody searches your brand name yet, push Meta to 60. If your search volume is healthy and your average order value clears 2,000 TL, move Google to 65. Rebalance every six weeks using cost per new customer, not the ROAS your dashboards report.

A home textiles brand put all 80,000 TL of its monthly budget into Meta last season and saw a 4.2 ROAS on the dashboard. Moving 30,000 TL into brand and category search lifted total revenue 34 percent while Meta's reported ROAS fell to 3.8. Anyone reading one platform report would have called that a mistake.

According to figures compiled by Coozmoo, Google took 39 percent of digital ad spend in 2023 and Meta 18 percent. Market share does not decide your split. It only proves neither platform has a scale problem, so the real question is the order your money gets spent in.

The ratios below assume a Turkish store spending 50,000 TL to 500,000 TL a month. Exchange rates, shipping and returns all eat margin here, so copying US ROAS targets one to one will mislead you.

Where Google Ads wins: the intent is already on the table

Someone typing "8 piece stainless steel cookware set price" has finished deciding. The need is defined and all that remains is picking a seller. Google Ads puts you in front of that person at that second, and it costs a lot, because six competitors want the same second.

Search and Shopping as the backbone

Per Dancing Chicken's 2024-2025 analysis, Google Ads converts at 3.75 percent on average and past 12 percent in automotive and healthcare. Shopping ads pull over 30 percent of sponsored clicks in online stores. Since image, price and brand name appear together, shoppers filter before reaching your site and wasted clicks drop. The same Coozmoo data puts 73 percent of US search ad clicks on phones.

Performance Max and the YouTube side

Google's 2025 year in review reports a 26 percent lift in conversions per dollar for Demand Gen campaigns through Performance Max. One campaign pools Search, YouTube, Display, Gmail and Discover inventory, and it works only when the product feed is clean. Missing titles, GTINs or category fields push it toward cheap inventory, where it cannibalises brand searches you would have won free.

Where the bill balloons

• Average cost per click sits near 2.69 dollars, above 6.50 in home services and 8.50 in legal.
• Keep brand keywords in a separate search campaign and add a brand exclusion list to Performance Max, or it buys traffic that was already yours and reports it as a win.
• Broad match plus smart bidding leaks money into irrelevant queries. Read the search terms report weekly.

For high ticket items buyers compare on specifications, Google is close to unarguable: appliances, electronics, spare parts, B2B services. A shopper comparing models who cannot find you simply buys elsewhere, and no report shows you that loss.

Where Meta Ads wins: you manufacture the demand

Nobody scrolling Instagram is shopping for cookware. Plenty of them will buy some after the right video. That makes Meta Ads a complement to search rather than a rival: sell a category with no search volume and you build demand here first, then collect it on Google.

Audience structure and the cost advantage

Coozmoo puts Meta's average conversion rate at 8.78 percent, above 10 percent in some verticals. It beats Google because it measures different behaviour: Meta conversions cluster around lower priced, quickly decided purchases. Search Engine Journal's 2025 report gives Facebook an all industry CPC of 1.88 dollars against Google's 4.66. Cheap traffic is not ready traffic, and the same person usually needs three or four exposures.

Creative volume is the real lever

What carries a Meta account is not your bidding strategy, it is how many videos you ship. Produce fewer than three fresh Reels formats a week and the audience fatigues, frequency climbs and CPM follows. One furniture brand ran a single studio shoot for three months and watched cost per order double after week four. The fix was not a bigger production. Customer unboxing clips, assembly footage and raw shots of the piece in a real living room pulled cost per order back down. Testing formats against each other in a proper split test beats raising the budget.

What Advantage+ takes over

As WordStream's 2025 trend analysis describes, Advantage+ campaigns hand targeting, placement and budget to the system. Give that up and creative plus your offer page are the only levers left. Launching them before cleaning audience lists and finishing the Conversions API setup asks an algorithm to decide well on bad data.

The numbers side by side: cost, conversion, return

Keep the platform choice out of the realm of opinion. The values below are industry averages, not facts about your account. Write your own figures into the same table at the end of month three.

Criterion	Google Ads	Meta Ads
Average cost per click	2.69 dollars, up to 8.50 by vertical	1.06 to 1.88 dollars
Average conversion rate	3.75 percent, up to 12 percent	8.78 percent average
Ecommerce ROAS range	4:1 to 6:1	6:1 to 7.5:1
Learning period	2 to 4 weeks	7 to 14 days
Main lever	Product feed and keyword structure	Creative volume and offer
Strongest moment	Demand already formed	Demand not yet formed

That conversion rate gap confuses people. Meta looks better on paper, yet a slice of those conversions come from shoppers who already knew you. Google looks worse, yet for most brands it brings in more genuinely new customers. Until you add both dashboards together and derive one blended cost per new customer, neither number settles the argument. Build the model with the ROAS formula and your gross margin in the same sheet, because a brand running on a 22 percent margin loses money at 4:1.

There is a second trap. Both platforms claim the same sale. A shopper sees your Instagram ad, searches your brand on Google the next day, then buys. Meta books it through view attribution, Google books it through last click. Add the two dashboards up and your reported revenue runs 20 to 40 percent above reality. The honest figure lives in your accounting records.

Spreading budget across the funnel

Choose a funnel stage before you choose a platform. The same 100,000 TL produces three different outcomes depending on which gap it fills.

Top of funnel: the unknown brand problem

Fewer than 500 monthly brand searches means there is nothing waiting for you on Google. Put 55 to 60 percent of the budget into broad Meta audiences and Reels formats. Judge it on three second view rate and new users landing on site, not on ROAS. Month one may return 1.5 and that is survivable.

Mid funnel: the comparison shopper

Now they know you and are weighing you against alternatives. Brand terms, category terms and Shopping carry the Google side. On Meta, catalogue ads chase everyone who viewed a product page without adding to cart. Split the budget roughly down the middle here. Traffic stops being cheap, but this is the closest volume to a purchase.

Bottom of funnel and repeat purchase

Cart abandoners, 30 day product viewers, one time buyers who never returned. Those three lists take 15 to 20 percent of total spend and usually deliver the strongest return. If abandonment runs above 70 percent, the problem sits in your checkout, not your ads. Fix the conversion rate leaks on the site before paying to push more traffic into a leaking bucket.

Loyalty is the stage most brands leave empty. Uploading your existing customer list to both platforms and announcing a new collection costs a fraction of acquiring someone cold, and the same list feeds your lookalike audiences.

Splits by vertical, with a worked example in TL

A universal ratio does not exist, but vertical ratios do. The table assumes 100,000 TL a month. Spending 40,000 TL changes the absolute numbers and nothing else.

Vertical	Google share	Meta share	On 100,000 TL	Why
Fashion and apparel	45 percent	55 percent	45,000 TL / 55,000 TL	Visual discovery triggers the purchase
Electronics and appliances	65 percent	35 percent	65,000 TL / 35,000 TL	Model and price comparison happens in search
Beauty and personal care	40 percent	60 percent	40,000 TL / 60,000 TL	User video does the persuading
Home and furniture	50 percent	50 percent	50,000 TL / 50,000 TL	Long decision cycle needs both touchpoints
Spare parts and industrial	75 percent	25 percent	75,000 TL / 25,000 TL	Demand exists only as part number queries
Food and supplements	35 percent	65 percent	35,000 TL / 65,000 TL	Subscription and repeat buying build socially

Work a real example. A beauty brand averages 1,400 TL per order at 35 percent gross margin and spends 100,000 TL a month. Each order leaves 490 TL, minus 90 TL for shipping and packaging, so 400 TL remains. Break even needs roughly 250 orders. If the 60,000 TL on Meta buys orders at 320 TL, that is 187. If the 40,000 TL on Google holds at 260 TL, another 153 arrive. Total 340, comfortably above break even.

Run the same maths with a 12 percent return rate and it shifts. Forty orders come back, shipping is paid twice, net margin barely covers the spend. Budget allocation is an accounting decision wearing a marketing costume. Keeping catalogue, stock and order data in one place through an ecommerce platform built for it produces this calculation faster than a monthly spreadsheet.

Constraints specific to Turkey: KVKK, IYS and marketplaces

Three things break when you copy tactics from US blogs.

KVKK and uploading customer lists

Uploading a customer list to Meta or Google transfers personal data abroad. If your consent text does not cover that transfer, the upload becomes a legal exposure rather than a growth tactic. Have a lawyer review the privacy notice on your signup form and split marketing consent from data transfer consent into two boxes. Building audiences from site traffic closes most of that risk.

IYS and what happens after the click

Any email address or phone number you collect through ads becomes a commercial message only after it is registered with the Message Management System. Texting a lead form number that never made it into IYS invites a fine. Budget the operational cost of that step alongside the media spend, and decide early whether you are collecting email consent, SMS consent or both, since they are gathered separately.

Marketplace spend versus your own store

Selling on Trendyol and Hepsiburada splits your budget three ways: marketplace ads, Google, Meta. Marketplace commission runs 15 to 22 percent while iyzico or PayTR on your own site costs around 2 percent. Your own domain carries visibly better margin, yet most brands pour everything into the marketplace because conversion there feels easy. Carve part of the Meta budget into store exclusive offers instead, and pair it with a free shipping threshold negotiated through Yurtici Kargo or Aras so your delivery promise stops losing. How brands build their own direct channel is worth reading before that shift.

Measurement setup and the first 30 days

A badly measured budget costs more than a badly allocated one. Complete the setup in this order.

Setup order

• In Google Ads, open Tools and Settings, go to Conversions, and define the purchase action with value tracking enabled.
• Turn on enhanced conversions and send a hashed email from the thank you page.
• Connect the Conversions API server side in Meta Events Manager and get Event Match Quality above 7.
• Switch GA4 attribution to data driven and fix the conversion window at 30 days.
• Standardise your UTM template across both platforms, otherwise one channel shows up twice in reporting under different spellings.

The weekly rhythm

Week one is verification. Place a test order and confirm it lands in all three dashboards. Week two, open the budget at your vertical's ratio and leave it alone, since daily meddling during the learning phase starves the algorithm. Week three, read the creative and search terms reports, then cut weak placements and irrelevant queries. Week four brings the first reallocation: move 10 percent toward whichever platform shows lower cost per new customer, no more.

If the metrics themselves feel tangled, seeing how CTR, CPC, CPM and CPA push against each other in one table clears it up quickly. Tomorrow morning the job is concrete: set the Google Ads search terms report to the last 30 days, add every term that spent over 500 TL without a sale to your negative list, then pause any Meta ad set with a frequency above 3 and upload two fresh creatives in its place.

Frequently Asked Questions

Should I run both platforms on 20,000 TL a month?

No. Split it and neither account finishes learning. Pick one and stay three months: Meta if nobody searches your brand, Google if your product already has query volume.

Performance Max or classic search campaigns?

Both, kept apart. Hold brand and high volume category terms in a dedicated search campaign, then let Performance Max take the rest. Without a brand exclusion list it buys traffic already headed your way and reports it as incremental.

Meta shows a 6 ROAS but revenue is flat. Why?

Most of the budget is probably sitting in retargeting, where the audience would have bought anyway. Check what share of spend reaches cold audiences. Below 40 percent, the account measures existing demand rather than creating customers.

Is Meta measurement still trustworthy after the iOS changes?

Largely yes, provided the Conversions API is live and match quality is high. Without it, reported numbers land below actual sales. Compare dashboard revenue against weekly accounting figures first.

How often should I change the split?

Every six weeks. More frequent intervention disrupts how bidding algorithms gather data. The exceptions are a broken tracking setup or a product going out of stock.

Which product feed field matters most for Shopping?

The title. Fit brand, model, colour, size and category into the first 70 characters. Missing GTINs suppress impressions badly, and stock status needs refreshing twice a day.

Can I reuse the same creative on both platforms?

A vertical video can run as a Reel and a YouTube Short, but the copy should differ. Meta needs a visual hook in the first three seconds, YouTube viewers give you longer. Square stays safe on both.

How much of the ad budget should move to SEO?

If ads are profitable, SEO is a separate line item, not a deduction from media spend. A balanced plan across organic and paid shows how the two feed each other over a year.

When should I switch off a campaign?

Kill an ad set once it hits three times your target cost per order with more than 50 clicks recorded. Less data than that and you usually kill a set that would have worked.

Does the split change during Black Friday?

Yes. Meta CPMs rise sharply through November and the same audience gets expensive. Finish warming audiences by mid October, then shift over half the November budget into retargeting and brand search.