Skip to main content
Metrics type: Cross-Platform MetricsCategory: Ad Platform

At a glance

The share of branded paid clicks on TikTok that would most likely have arrived free through organic search or direct channels anyway. When a TikTok campaign targets people already looking for your brand, you are often paying to re-acquire demand that Google organic, branded search, or your own direct traffic was already going to capture for free. That is pure margin reclaim: cap or exclude the warm brand-affinity audience and keep the revenue without the ad cost. Caveat: this is an estimate of incrementality, not a measured experiment. The only definitive proof is a holdout or geo lift test; this card is the directional flag that tells you when one is worth running.

Calculation

Calculated automatically from your TikTok Ads data joined to your organic search and direct traffic baseline. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

Worked example

A US homeware brand running a TikTok brand-affinity and follower-retargeting campaign alongside healthy Google organic and email. The 30-day window is 13 Feb 26 to 14 Mar 26. All figures are illustrative.
  1. The gauge reads about 38% across branded paid clicks, above the 30% threshold, so it fires. The two warm audiences are doing most of the cannibalising.
  2. The follower / existing-customer retargeting at 9.1x ROAS is the most misleading line. Those people were the most likely to return on their own through Google branded search, email, or direct, so an estimated 55% of those clicks were probably free. The headline ROAS flatters a campaign that is largely defending revenue you would have kept anyway.
  3. Cold prospecting is clean. The lookalike and interest audiences sit near 5% estimated free share, because those users had no prior brand intent. That spend is genuinely incremental, which is exactly what TikTok is good at (discovery-time demand creation). Leave it running.
  4. The reclaim action is targeted, not blunt. Rather than killing all branded spend, cap or exclude the follower and recent-brand-engager clusters, then watch whether Google organic and direct revenue holds. If total revenue is flat after the cut, the spend was indeed cannibalising.
  5. Prove it before scaling the decision. A two-week holdout (suppress the warm audience for half the eligible users) measures the true lift. The card points you at the test; the test gives you the defensible number.
Quick reads:
  • High estimated free share + high reported ROAS + warm audience = classic cannibalisation. Test a holdout.
  • Low estimated free share + cold audience = incremental discovery spend. Keep it; it is what TikTok does best.
  • Estimated free share rising over time = organic search and email maturing; the branded TikTok campaign is increasingly redundant.

Sibling cards merchants should reference together

Reconciling against TikTok Ads Manager

Where to look in TikTok Ads Manager: TikTok Ads Manager does not surface a cannibalisation metric. The honest in-platform proxies are:
  • TikTok Ads Manager > Experiments / Lift study, which can run a holdout to measure incremental conversions. This is the definitive method and the one this card is steering you toward.
  • The audience definitions behind each ad group tell you which campaigns target warm versus cold users; a warm audience is the prerequisite for cannibalisation.
Other views that look related but are not:
  • Reported ROAS by campaign: a high warm-audience ROAS looks like success but is exactly the figure cannibalisation inflates. It is the symptom, not the diagnosis.
  • Attribution settings: changing the window does not reveal incrementality; only a lift test does.
Why our number may legitimately differ from TikTok: Cross-connector reconciliation:

Known limitations / merchant FAQs

Is this a measured number or an estimate? An estimate. It models how many branded paid clicks would likely have arrived free through Google organic search or direct channels, based on observed behaviour. It is not a measured incrementality result. The only definitive measure is a holdout or geo lift test, which this card is designed to prompt. Why do my highest-ROAS campaigns get flagged? Because warm audiences, brand engagers, followers, recent visitors, and existing customers, post the highest reported ROAS precisely because they were already going to convert. High ROAS on a warm audience is the signature of cannibalisation, not of genius targeting. Should I just turn off all branded campaigns? No, not bluntly. Some branded spend genuinely defends against competitors and accelerates a return that would otherwise be slow. Cut surgically: cap or exclude the warmest clusters first, watch whether Google organic and direct revenue holds, and keep the spend that proves incremental. How is this different from branded keyword cannibalisation on Google Ads? The mechanism is the same idea, paying for demand you would have won for free, but the lever differs. On Google you exclude or down-bid branded keywords; on TikTok you exclude or cap warm brand-affinity and follower audiences. The diagnostic question is identical: would this click have come for free through organic search? Why a 30-day window? Organic search and email behaviour is noisy week to week, and a single promotion can distort a short window. Thirty days gives a stable read of the steady-state overlap while still reacting within a month to a real shift. What do I do once it fires? Treat it as a test prompt. Set up a holdout for the flagged warm audience for two weeks, then compare total revenue and organic search and direct revenue against the exposed period. If totals hold with the spend removed, reallocate that budget to cold prospecting (where TikTok genuinely earns its keep) or pause it.

Tracked live in Vortex IQ Nerve Centre

Branded Paid Clicks Cannibalising Organic Search is one of hundreds of KPI pulses Vortex IQ tracks across TikTok Ads and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.