Every ad platform will tell you it's the one your store can't afford to miss. Spread a modest budget across all of them, though, and none gets enough to work properly. For most online stores, the better plan is to go deep on two platforms before adding a third.
The short answer: start with Google and Meta. Google reaches people who are already searching for what you sell, and Meta finds people who don't know they want it yet. Between them they cover most of the buying journey. Add TikTok once both are working, and only if your products show well on video and you can keep up with the creative it demands.
What each platform does best
| Platform | Where it reaches people | Best for | Watch out for |
|---|---|---|---|
| Search, Shopping, YouTube, Gmail, and partner sites | Capturing people already looking for what you sell | Automated campaigns can spend heavily on your own brand name | |
| Meta | Facebook and Instagram feeds, Stories, and Reels | Creating demand among people who haven't found you yet | Results depend heavily on a steady supply of good creative |
| TikTok | Short-form video in the For You feed | Products that demonstrate well and can catch on | Lower purchase intent, heavy creative demands, and harder measurement |
Why Google and Meta come first
- Between them, they cover both kinds of shopper. On Google, people tell you what they want by searching for it. On Meta, you reach them before they've started looking. A store that runs both gets demand from Meta and captures it on Google, and you'll often see branded searches rise as Meta campaigns scale.
- Their automation needs data, and splitting the budget starves it. Google's Performance Max and Meta's Advantage+ campaigns decide who sees your ads based on the conversions they've seen. Meta's learning phase, for example, is built around roughly 50 results a week. Split the same budget across four platforms and none of them gets enough results to learn.
- Measurement stays manageable. Every platform reports its own results, and they overlap. Two platforms are easy to reconcile against your actual sales. Four is a spreadsheet project every month.
Getting the most from Google
- Fix the product feed first. Shopping ads and Performance Max build your ads from your product feed, so titles, images, prices, and availability all affect how often you show up. Clear, specific product titles tend to help more than anything else.
- Watch what Performance Max spends on your brand. It can put a lot of budget toward people searching for your store by name, who would probably have found you anyway. Keep an eye on the split, and run branded search separately so you can see it.
- Reward new customers. Google Ads can bid more for people who haven't bought from you before. It's worth turning on, and worth measuring with new customer acquisition cost rather than ROAS alone.
Getting the most from Meta
- Let the targeting stay broad. Meta's automated targeting now usually outperforms carefully built interest audiences and lookalikes, as long as it has conversion data to learn from.
- Treat creative as your targeting. With broad targeting, the ad itself decides who responds, so the variety of your creative matters more than any audience setting. Our guide to Meta Ads creative covers what to make.
- Run catalog ads. If you sell more than a handful of products, ads that pull straight from your catalog and show people what they've viewed are usually among the most efficient campaigns in the account.
When TikTok makes sense
TikTok can work very well for the right store. It's worth testing if most of these are true:
- Your product demonstrates well on video. Visible results, a satisfying use, or something people haven't seen before.
- You can make a lot of video. TikTok ads wear out quickly, so you'll need new creative every week or two, often from creators rather than a studio shoot.
- Your margins can absorb a test. Creator fees and production add to the cost, and it can take a few months to find what works.
- Your customers are there. TikTok's audience is broad now, but it still skews younger than Facebook's.
- Google and Meta are already profitable. TikTok is for growth on top of a working base, not for fixing one that isn't working.
The limitations of TikTok ads
- Lower purchase intent. People are there to be entertained, so many will watch, enjoy, and move on. Expect lower conversion rates than search, especially at first.
- A constant need for new creative. What works on TikTok looks like TikTok, not like an ad, and the same video rarely lasts long.
- Harder measurement. Many TikTok sales happen later, after someone searches for you on Google or comes back directly. TikTok's own reports will credit some of those, and other tools won't credit any, so the truth is usually somewhere in between.
How to split the budget
There's no universal formula, but this is where we usually start:
- Start with Google and Meta only. Split between them based on how people find products like yours. If people search for what you sell, lean toward Google. If it's something people need to see before they want it, lean toward Meta.
- Stay there until both are steady and profitable. That means consistent results month to month, measured on new customer acquisition cost, not just each platform's own ROAS.
- Then set aside around 10 to 20% to test a new channel. TikTok is a common choice, and Pinterest, YouTube, or Amazon can make more sense depending on the product. Keep the test running for at least a couple of months before judging it.
- Judge the whole account, not each platform. Watch your total new customers and your overall cost to acquire them. If adding TikTok lifts total sales, it's working, even if its own reports look modest.
Go deep on two platforms before you go wide on four.
The bottom line: Google captures the demand that already exists, Meta creates more of it, and together they're the right foundation for almost every online store. TikTok is a strong third channel for products that shine on video, once the first two are working. If you'd like a second opinion on how your budget is split, we're happy to take a look.