Facebook ads vs Google ads: where should a startup start?
A startup should start with Google ads if people already search for its product or category, and with Facebook ads if they don't. Google Search and Shopping capture existing demand, while Meta creates new demand by reaching people who weren't looking. Prove one channel first, then add the second and judge both together with MER, not platform ROAS.
Start with Google ads if people already search for your product or category, and start with Facebook ads (Meta) if they don't. Google Search and Shopping capture demand that exists. Meta creates demand by showing your product to people who weren't looking. Most new-category products start on Meta. Most products people already shop for start on Google. Add the second channel once the first one pays its way.
What is the difference between Facebook ads and Google ads?
Facebook ads interrupt people, and Google Search ads answer them. On Meta (Facebook and Instagram), your ad appears while someone scrolls, so it has to make them want something they weren't thinking about. That's demand creation. On Google Search and Shopping, your ad appears after someone types a query like "organic dog shampoo," so they already want the thing. That's demand capture.
Google also sells demand creation through YouTube, Display and Demand Gen. Performance Max serves across all of it plus Search and Shopping, and Google's AI picks placements from your goals, assets, feed and budget. For a first channel, the real comparison is Meta versus Google Search and Shopping.
How do you decide which one to start with?
Estimate whether enough people search for what you sell. In Google's Keyword Planner, add up monthly searches for your product, your category, the problem it solves and the products it replaces. Low volume for your exact product name doesn't mean no demand.
Say all those terms add up to 300 searches a month. WordStream's 2026 search benchmarks put the average click-through rate at 6.64% and the average conversion rate at 8.18%. Even if your ad showed on every search, average rates give about 20 clicks and fewer than 2 conversions a month. That's too little to learn from. Start on Meta.
If the category gets tens of thousands of searches a month, buyers are already looking. Start on Google, as long as the clicks fit your margins.
| Your situation | Where to start |
|---|---|
| Established category, lots of buying searches | Google Search or Shopping |
| New or visual product that needs a demo | Meta, with video creative |
| High search volume but very expensive clicks | Check break-even cost per purchase before committing to Google |
| People search your brand name, few search the category | Keep branded search small and separate; find new customers on Meta |
| Thin margins or a weak offer | Fix price, bundles or the site before scaling either channel |
How do Meta ads and Google ads compare?
| Meta (Facebook and Instagram) | Google (Search, Shopping, PMax) | |
|---|---|---|
| How targeting works | Meta's algorithm finds likely buyers from your creative, pixel data and broad audiences | Keywords and queries on Search; product feed matching on Shopping |
| Scale limit | Creative and audience: fresh ads keep reach growing | Search volume: you can't buy searches that don't happen |
| Creative needs | Heavy: new videos and images often, because ads wear out | Light for Search and Shopping (text ads, a clean feed); heavier for PMax and YouTube |
| Learning considerations | An ad set typically needs about 50 optimization events in 7 days to exit learning | Google advises running new Performance Max campaigns for at least 6 weeks |
| Benchmark cost | Median CPA $38.99, median ROAS 1.88, median CPM $15.06 | Search: average CPC $5.42, conversion rate 8.18%, cost per lead $66.69 |
| Attribution defaults | Standard setting for website conversions: 7-day click-through, 1-day engage-through, 1-day view-through | Model-based credit across Google ad clicks for most conversions; 30-day click-through window |
Meta figures come from Triple Whale's benchmarks on over 40,000 ecommerce brands that use its platform (August 2025 to July 2026). Google figures come from WordStream's 2026 benchmarks: 13,474 US search campaigns on Google and Microsoft Ads (April 2025 to March 2026). WordStream's averages are medians, and its leads include calls and form fills. Neither learning figure is a minimum budget.
What do Facebook ads and Google ads cost for a startup?
Both cost more per customer than most founders expect, and the benchmarks measure different things. Meta's median cost per purchase was $38.99 in Triple Whale's ecommerce data. WordStream's 2026 search data puts the median cost per lead at $49.40 for shopping, collectibles and gifts, with a $4.14 cost per click. Those leads include calls and form fills, so cost per purchase would likely run higher.
The useful comparison is against your own break-even cost per purchase: average order value times contribution margin, where the margin is what's left of a first order after product cost, shipping, payment fees, discounts and returns, but before ad spend. If your break-even is $30 and benchmarks run near $40, fix the economics first. Known repeat purchases can justify paying more. How much to budget for each platform is covered in how much a startup should spend on ads.
Which channel fits a new-category product vs a commodity?
A new-category product usually belongs on Meta first. Imagine launching an unfamiliar countertop food-waste appliance. If few buyers know what to call it, search volume stays small, and a 15-second video of scraps turning into dry grounds has to introduce it. If it works, some viewers later search the brand name, and branded search becomes worth buying.
A commodity product usually belongs on Google first. Take a set of bath towels. People search "cotton bath towel set" every day and compare prices on Shopping. Here's how the math works for a hypothetical brand, assuming the 50% margin already covers product, shipping, fees and returns:
| Before | After adding a bundle | |
|---|---|---|
| Average order value | $90 | $120 |
| Contribution margin before ads | 50% | 50% |
| Break-even cost per purchase | $45 | $60 |
| Shopping cost per click (assumed) | $1.50 | $1.50 |
| Conversion rate (assumed) | 3% | 3% |
| Cost per purchase ($1.50 ÷ 3%) | $50 | $50 |
| Contribution per first order after ads | −$5 | +$10 |
On a commodity, the auction sets the ad cost, so margin and order value decide whether Google works.
Why do Meta and Google report different results?
Each platform counts sales by its own rules, so both can claim the same sale. Meta's standard setting credits a purchase within 7 days of a link click, or within 1 day of a view or other engagement. Google's default model for most conversions splits credit across Google ad interactions, within a 30-day click window. A shopper who clicks your Instagram ad, then clicks a branded Google ad two days later and buys, can be counted by both.
So the two platforms' reported revenue can add up to more than your store made, and branded search can catch buyers another channel brought in. Which platform gets too much credit depends on your customers and settings. Don't settle Meta vs Google on platform ROAS; compare each to your break-even, as explained in what a good ROAS is.
When should a startup run both, and how do you judge them?
Run both once the first channel is profitable at a steady budget and you can fund a real test of the second. Track MER: total store revenue divided by total marketing spend. MER ignores platform attribution. Fees are left out below for simplicity; include them in your real MER.
| Month 1: Google only | Month 2: Google + Meta | |
|---|---|---|
| Ad spend | $4,000 | $8,000 |
| Revenue reported by platforms | $17,000 | $30,000 ($18,000 Google, $12,000 Meta) |
| Actual store revenue | $16,000 | $26,000 |
| MER | 4.0 | 3.25 |
MER fell, but revenue rose $10,000 while spend rose $4,000. At a 50% contribution margin, that extra revenue leaves $5,000, or $1,000 more than the added spend. A before-and-after view can't prove Meta caused the lift, though. Seasonality, a promotion or organic growth could explain some of it. To measure what Meta added, run a holdout or geo test that pauses ads for some people or regions. Read MER next to contribution profit and new-customer cost; see MER vs ROAS for the full method.
Wadabu manages up to two ad platforms for $3,000 a month or 10% of ad spend, whichever is greater (ad spend paid separately), against one MER target.
Final thoughts
Start where your buyers already are: on Google if they search for your product, on Meta if you have to show it to them first. Prove one channel, then add the second and judge both by MER and contribution profit. If you want a team to run that test for you, see Wadabu's paid ads service for startups.
Frequently asked questions
How much should a startup budget to test Facebook ads vs Google ads?
A Meta test needs enough budget for about 50 optimization events per ad set in 7 days, so at Triple Whale's $38.99 median CPA, roughly $1,950 a week for one ad set. A Google Search test scales with click costs and search volume, so it can often start smaller. Run one channel first.
Are Facebook ads cheaper than Google ads?
Facebook ads and Google ads are hard to compare on price alone. Triple Whale put the median Meta cost per purchase at $38.99 for August 2025 to July 2026. WordStream's 2026 search data puts the average Google search click at $5.42 and cost per lead at $66.69. Compare each against the brand's own break-even cost per purchase.
Can a startup run Meta ads and Google ads at the same time?
A startup can run both once one channel is profitable and the budget can fund a real test of the second. Splitting a small budget across both from day one usually leaves each short of data. Track the combined result with MER and contribution profit, and use a holdout test to measure lift.
Why does Google Ads sometimes look better than Meta ads in reports?
Google Search often catches shoppers right before they buy, including branded searches from people who first saw a Meta ad. Each platform credits sales by its own attribution rules, so both can claim the same order. Which one looks inflated depends on how customers buy, so store revenue and MER make a fairer comparison.
Meta CPA, ROAS and CPM: Triple Whale Meta ads benchmarks (40,000+ ecommerce brands, August 2025 to July 2026, updated August 2026). Google CPC, CTR, conversion rate and cost per lead: WordStream by LocaliQ 2026 Google Ads benchmarks (13,474 US search campaigns, April 2025 to March 2026). Google attribution and Performance Max guidance: Google Ads Help. Meta attribution and learning phase: Jon Loomer Digital (2025–2026).
Written by Andrew Zam, Founder of Wadabu, a growth marketing agency for consumer-brand startups.
Want a plan for your brand?
Tell us where you are and where you want to be. We'll set up a quick call and map out your next moves.