TikTok is adopting a new form of ad measurement after claiming it deserved more credit for sales that come from its platform. It will now use a “self-attributing network” to report its successes.

The move puts TikTok more in line with how social media companies such as Meta have always measured. But for advertisers, a self-attributing network raises concerns that TikTok could become more of a closed system that grades its own homework, according to marketing experts.

“TikTok wants to hold their data closer to themselves,” said a tech expert at one of TikTok’s measurement partners, a firm that works with social media apps to track ad results for brands.

TikTok announced earlier this week that it was opening the “self-attributing network,” with plans to fully implement it by March 2024. The measurement reporting still runs through a group of approved third-party providers, including Kochava, AppsFlyer, Singular and Adjust, and TikTok still works with these measurement companies, but it will provide important conversion data through the self-attributing network. Conversions are when an ad leads to a result—a download or a sale. TikTok’s older measurement partner program could effectively track “last-click attribution,” when an ad on TikTok was the final touchpoint for the consumer.

The self-attributing network will report “view through” data, which tracks when ads are seen on TikTok. This measures when views contributed to a conversion, even when TikTok didn’t get that last click.

TikTok is launching the self-attributing network because, according to the company, 79% of all conversions attributed to TikTok by users themselves were missed by last-click attribution models.

A self-attributing network “measures its results internally,” according to measurement partner Singular’s website. Facebook, now Meta, first trotted out a self-attributing network more than a decade ago, according to measurement partners. Now, most social media apps have followed suit. TikTok sees its self-attributing network as a way to gain parity with rivals, taking credit for the efficacy of its ads.

A self-attributing network “will enable our platform to track and more accurately attribute app conversions, giving brands more powerful insight into TikTok's contribution to app campaigns,” TikTok said in its announcement.

Also read: TikTok opens search ads to all advertisers

What this means for advertisers is that they will start to see higher conversion rates on ads on TikTok. But self-attribution also brings trust issues. Digital advertisers have deep-seated skepticism of online measurement schemes, and even more so when that data is restricted by the platforms. TV advertisers are used to demanding stricter scrutiny of metrics such as TV ratings, which go through layers of oversight and independent research through third parties such as Nielsen, to properly value the ad inventory.

TikTok’s new measurement system comes amid a broader shift online, with changes to how data gets shared among apps, devices and web browsers. Apple limited the flow of data from iPhones and Safari web browsers, and marketers can no longer granularly track when an ad on apps lead to conversions.

Apple’s data crackdown sent the app world into a tizzy, looking for fixes to conversion reporting. Not to mention, Apple’s answer to measurement, its SKAdNetwork, has been glitchy. Just last month, the latest iteration of Apple’s SKAdNetwork had a platform-wide bug that stopped reporting conversions to measurement partners, according to a blog from measurement firm Adjust. Google, meanwhile, is making similar changes as Apple to limit data sharing when it rolls out “Privacy Sandbox” tools more widely next year; this includes its plan to kill cookies, which track online behavior, on Chrome.

While the device makers and browsers make these changes, the apps, including TikTok, are tinkering with how they share data, too. The apps are developing conversion APIs, data clean rooms and direct integrations with major marketers, through tools such as tracking pixels, to share first-party data to measure ad performance. Meta has been working on what it calls “privacy-enhanced technologies,” using encrypted signals and aggregated data sets that are more anonymous to report to marketers. But marketers want ways to audit these new forms of measurement to prove the accuracy of the results they are handed.

TikTok’s relationship with advertisers also is complicated because of its ties to its Chinese-based owner ByteDance. TikTok has been under a microscope in the U.S. because of its influence in social media and its foreign ownership. In some cases, advertisers have been slower to adopt some tools, such as tracking pixels, which marketers put on their websites to measure when ads from apps sent consumers their way. 

“TikTok being a self-attributing network isn’t surprising,” another executive at a measurement firm said. “Them being a Chinese company and a self-attributing network could be the part that raises eyebrows.”

TikTok’s move to a self-attributing network “increases privacy, but it also decreases transparency for advertisers,” said another executive at a TikTok measurement firm. The measurement partners won’t get access to “user-level data for view-through campaigns,” this person said.

Marketers still need the measurement partners, however, as part of a multipronged approach to measurement, so the partners don’t feel they are being squeezed out, according to Charles Manning, CEO of Kochava, another measurement partner.

“A lot of people will be really negative about a self-attributing network, because it allows the publisher to grade their own homework,” Manning said. “But any marketing measurement partner worth their weight is cross-comparing multiple self-attributing networks to validate ad performance.”

The digital advertising world is moving to more forms of aggregated data for privacy reasons, Manning said. “The pendulum is swinging from advertisers having all the insights to premium publishers having control over what insights are available and all of that is over privacy reasons.”