TikTok Ad Network is now in US campaigns. Decide the placement before you build

KEY TAKEAWAYS

Placement locks when the ad group is created. Taking the network out later means building a new ad group.
Smart+ defaults to automatic placement, the opt-out goes through a rep, and the block list is off until you switch it on in each ad group.
In the US the network serves interstitial and rewarded ads only. A rewarded ad pays the viewer to watch to the end, so read engaged-view credit with care.
Break-even match rate equals network reported cost divided by feed reported cost, times the feed's match rate. In our hypothetical example it is 45%.

On October 5, 2026, TikTok opened its off-platform ad network to campaigns aimed at US users. Most coverage treats it as a switch you can try and turn off. It is not. TikTok's help centre says an ad group's placement is chosen at creation and cannot be changed afterwards. So the decision belongs before the build, and it needs a number. This article gives the rules as documented, a sum for judging the network, a decision table and a method.

What changed on October 5, 2026?

TikTok Ad Network is the new advertiser-facing name for Pangle, the network ByteDance launched in 2018. TikTok's help centre puts it at nearly 400,000 apps and publishers, with more than 1 billion daily active users across more than 40 genres. Marketing Dive and Digiday reported the US opening on October 5, from Advertising Week New York. Music Business Worldwide and The Keyword both describe the US as the network's 48th market.

It is not a separate product. It is a placement inside TikTok Ads Manager, next to TikTok itself. Some help pages still call it TikTok Pangle, so expect either name in the interface.

Is the network on by default?

It depends on the campaign type. The help pages give four rules.

Smart+ starts on automatic placement. One help page says automatic placement is the default for Smart+ campaigns. Another says that, depending on region, automatic placement may include partner apps through the network.
The Smart+ opt-out goes through a person. The page tells advertisers who do not want automatic placement to contact their sales representative or customer support.
Manual campaigns let you choose. The manual options listed are TikTok, Global App Bundle and the network.
The choice is permanent for that ad group. The help pages repeat, in the same words each time, that placement is set at creation and cannot be changed later.

The controls are thinner than the launch language suggests. A block list holds up to 10,000 app bundle IDs. It does nothing until you switch on a toggle in the advanced settings of each ad group, and TikTok warns that blocking can reduce impressions and performance. Before launch you can see a top 100 publisher list. After launch a delivery report shows where ads ran, with a lag of 1 to 2 days, and TikTok notes it may leave out some publishers for contractual reasons. The pages we read describe no way to limit delivery to an approved list of apps. A media buyer quoted by Digiday asked for exactly that.

What does a US impression on the network look like?

Not like a feed video. TikTok's format page lists six formats for the network, then states that the US market currently gets two: interstitial ads and rewarded ads.

An interstitial is a full-screen ad shown at a break in someone else's app, such as the end of a game level. A rewarded ad gives the viewer an in-app reward for watching the whole ad.

That second format matters for measurement. TikTok counts an engaged view when someone watches at least 6 seconds of an ad, and engaged-view attribution applies to Sales, Lead generation and App promotion objectives. Its help page lists windows of off, 1, 7 or 28 days. An older page on attribution windows lists only 1 or 7 days for engaged views, 1 or 7 for clicks, and off or 1 day for plain views.

Here is our inference, and it is only that. A person paid in game currency to finish an ad will pass 6 seconds almost every time. Neither attribution page says network inventory is treated differently. So a rewarded impression can earn engaged-view credit for a sale the viewer was going to make anyway. We made the same argument about low-bar views in the piece on Demand Gen view-through conversions.

Do TikTok's 65% and 63% figures apply to you?

The launch came with strong numbers. As Music Business Worldwide and The Keyword report them, campaigns on automatic placement saw a 65% better cost per acquisition and a 63% better return on ad spend than single-placement campaigns, with 14% higher budget utilisation. TikTok also says half the network's users are not on TikTok, which it sizes at 95 million people, and that people on both see 31% more touchpoints.

Check the origin before using them. The cost and return figures come from an internal test in April 2024 across Pangle's markets at the time. The US was not one of them. The reports give no sample size. The 95 million is internal US measurement. None of it has been checked by an outside party, and all of it is counted with TikTok's own attribution.

Advice is split. Common Thread Collective suggests enabling the network on your best awareness and retargeting campaigns and watching cost per thousand, frequency and return for seven days. Mintec says keep it off for always-on campaigns and test only where the objective and measurement allow. TikAdSuite advises TikTok only for conversions and leads. None of the three publishes data.

The most useful evidence is from another platform. Jon Loomer tested Meta's Audience Network and found that when he optimised for link clicks, nearly all the budget went to the network. When he optimised for a deeper quality event, none did. He also found rewarded video inflating completed-view counts.

General inventory quality is a fair worry but not a measured one here. Pixalate's Q2 2026 benchmark, built on more than 80 billion programmatic impressions, mostly from open auctions, put invalid traffic at 41% for mobile apps against 20% for web. That is the open market, not TikTok's network. TikTok says it filters invalid traffic with automated and manual methods, and DoubleVerify and Integral Ad Science measure the network after the bid.

Our reading: both camps are partly right, and the deciding variable is the event you optimise to. If the event is one you can verify in your own records, the network can be tested fairly. If it is a click, a view or an unqualified form fill, cheap inventory will win budget for the wrong reason. One Crossmedia buyer told Digiday, "Having a larger ad network doesn't make us move more money into TikTok." Reach is not the test. Verified cost is.

How much of the network's result must be real?

Use three lines.

Verified cost per result = reported cost per result ÷ match rate. The match rate is the share of platform-reported conversions you can find in your own orders or CRM.
Break-even match rate for the network = (network reported cost ÷ feed reported cost) × feed match rate.
Keep the network only if its measured match rate is above that break-even.

A hypothetical example, with invented numbers. A brand spends $30,000 in a month across two ad groups.

	TikTok only	Network only
Spend	$20,000	$10,000
Reported conversions	400	400
Reported cost per conversion	$50	$25
Match rate in the order system	90%	to be measured

By our arithmetic the blended reported cost is $37.50, which looks like a clear win for adding the network. The feed's verified cost is $50 ÷ 0.90, or $55.56. The break-even match rate for the network is ($25 ÷ $50) × 0.90, which is 45%.

If 40% of network conversions match, that is 160 real orders at $62.50 each. Worse than the feed. If 60% match, that is 240 orders at $41.67. Better. The dashboard shows the same $25 in both cases.

Two cautions. Match on click-through conversions, because a view leaves no click ID to reconcile. And a match proves the order exists, not that the ad caused it. For that you need a holdout. TikTok's Conversion Lift Study is a managed service arranged through an account representative for eligible accounts, and its help page does not mention the network. We covered turning one experiment into a reporting factor in the article on Meta incremental attribution.

Which placement for which campaign?

Campaign	Placement at creation	Why
Smart+ for sales or leads, US	Ask your rep about opting out, or build a manual campaign with TikTok only	Automatic is the default and cannot be undone for that ad group
Manual, website purchases, tight cost target	TikTok only, plus a separate network-only test ad group	Keeps the comparison clean and the main ad group untouched
Lead generation judged on qualified leads	TikTok only until a match rate exists	Form fills are a shallow event, and Loomer found a shallow event sent his budget to network inventory
App installs, games or utilities	Automatic placement or a network-only test, judged on retained users	In-app inventory suits app offers. On iOS, TikTok notes automatic placement draws from the dedicated campaign quota for both placements
Reach or video views	TikTok only	Rewarded completions can flatter view metrics, by analogy with Loomer's Meta finding
Retargeting	Small network-only test, read on clicks	A buyer told Digiday this could be a real use. More touchpoints also means more frequency

A seven-step method

List every live ad group and its placement. Flag Smart+ campaigns and anything on automatic placement.
Decide before you build. Write the placement into the campaign brief. A wrong choice costs a new ad group and a fresh start for delivery.
Switch on the block list in every ad group that can reach the network. Start from the top 100 publisher list and remove apps you would not want next to your brand.
Build the test as its own ad group. Network only, with a distinct tracking tag. TikTok's performance page lists breakdowns by day, week and month only, and its iOS guidance says placement breakdowns under automatic placement may vary. Separate ad groups avoid relying on a split you may not get.
Set attribution deliberately. Use the same click window in both ad groups. Add custom columns that separate click, engaged-view and view conversions.
Reconcile. After enough conversions to read, match click-through conversions to orders or qualified leads and compute both verified costs. Our piece on sizing a creative test explains why a handful of conversions proves nothing.
Pull the delivery report weekly. Add poor apps to the block list. Re-run the sum each month.

Where this stops

We could not open TikTok's own announcement post during this research, so its figures are given as other outlets reported them. The help pages are inconsistent on naming and on engaged-view windows, and we have said where. Whether engaged-view credit applies to rewarded ads in the way we describe is undocumented. The network is four days old in the US, so no independent US performance data exists yet. The worked example is invented.

What holds regardless is the order of work. The placement is fixed at creation, so the test design comes first and the dashboard second.

If you want paid results judged against what your own records show, the waitlist is open.

Questions

What is TikTok Ad Network?

It is TikTok's off-platform ad network, known as Pangle until this year. TikTok's help centre says it covers nearly 400,000 apps and publishers with more than 1 billion daily active users. It opened to US-targeted campaigns on October 5, 2026 and is bought as a placement inside TikTok Ads Manager.

Is TikTok Ad Network turned on by default?

For Smart+ campaigns, automatic placement is the default, and TikTok's documentation says it may include partner apps through the network depending on region. To opt out you contact your sales representative or support. In a manual campaign you pick placements yourself when you create the ad group.

Can I remove TikTok Ad Network from an ad group that is already running?

No. TikTok's help centre says the placement is selected when the ad group is created and cannot be changed after that. To remove the network you build a new ad group with TikTok as the only placement. You can still switch on the block list in an existing ad group to exclude specific apps.

How do I know whether TikTok Ad Network conversions are real?

Run the network in its own ad group with its own tracking tag, then match its click-through conversions to orders or qualified leads in your own system. Divide the reported cost per result by the share that matches. Compare that verified cost with the same figure for a TikTok-only ad group.

Sources

TikTok Business Help Center, About TikTok Ad Network (formerly TikTok Pangle) ads.tiktok.com
TikTok Business Help Center, Automatic and Select Placement ads.tiktok.com
TikTok Business Help Center, About placement for your upgraded Smart+ experience ads.tiktok.com
TikTok Business Help Center, About eligible placements ads.tiktok.com
TikTok Business Help Center, Available ad formats for TikTok Ad Network ads.tiktok.com
TikTok Business Help Center, How to use the TikTok Pangle block list ads.tiktok.com
TikTok Business Help Center, About TikTok Pangle media transparency solutions ads.tiktok.com
TikTok Business Help Center, About TikTok Pangle's trust framework ads.tiktok.com
TikTok Business Help Center, About Engaged View-Through Attribution ads.tiktok.com
TikTok Business Help Center, About the attribution window on TikTok Ads Manager ads.tiktok.com
TikTok Business Help Center, About Conversion Lift Study ads.tiktok.com
TikTok Business Help Center, Best practices for iOS 14.5+ dedicated campaigns on the TikTok Ad Network ads.tiktok.com
TikTok Business Help Center, How to see your ads performance in TikTok Ads Manager ads.tiktok.com
Marketing Dive, TikTok broadens agentic AI play, brings global ad network to US marketingdive.com
Digiday, TikTok's ad network bet. More reach, more budget digiday.com
Music Business Worldwide, TikTok brings its global ad network to the US and unveils agentic AI tools musicbusinessworldwide.com
The Keyword, TikTok Ad Network off platform thekeyword.co
The Keyword, TikTok Ad Network Has to Prove Its Reach to Agency Buyers thekeyword.co
B&T, TikTok opens ad network to Aussie advertisers targeting US users bandt.com.au
Common Thread Collective, TikTok Advertising Week 2026 commonthreadco.com
Mintec, TikTok Now Sells Ads in 400,000 Other Apps. Should Your Budget Leave TikTok? mintec.co
Socialday, TikTok opens 400,000-app ad network to US advertisers, but Smart+ campaigns pull it in by default socialday.live
TikAdSuite, TikTok Ad Placement Options tikadsuite.com
Jon Loomer Digital, Should Meta Advertisers Ever Use the Audience Network Placement? jonloomer.com
Pixalate, Q2 2026 Ad Fraud Benchmarks Report for EMEA pixalate.com

Esh

Founder, Ethos Engine. Builds the Digital Authority Score and runs the placement marketplace. esh@ethosengine.io

NEXT STEP

See where your brand stands.

Your Digital Authority Score, who AI recommends for your target questions, and the placements that close each gap. Your first report is free.

Get your free report →