Deinfluencing passed its second year as a named TikTok trend in early 2025, and the budget collapse some predicted never arrived. Emarketer forecast in March 2025 that US influencer marketing spending would pass $10 billion that year, one year earlier than its prior call. What the trend changed sits elsewhere: in testing budgets, creative vetting, and how engagement gets read.
Skeptics called deinfluencing a passing mood of a bad economy. Two years of numbers argue for a longer view. The drift of Gen Z searches toward social platforms feeds the same demand for proof over polish, and the trends desk tracks that wider shift. Below is the deinfluencing trend data that now exists, and the holes still in it.
Where The Trend Started
The #deinfluencing hashtag took off on TikTok in early 2023. Swayable's study, published in May 2024, dates the rise to that year, as creators began telling followers what not to buy. The format spread fast because it flipped a familiar script. Instead of pitching a product, a creator panned one, often a viral item the audience already owned or had queued in a cart.
The timing mattered. Inflation had made shoppers suspicious of polished sponsorship, and a creator saying skip it read as honesty rather than performance. That framing did the real work. The format gave viewers permission to keep scrolling past a purchase.
Did Creator Budgets Fall
No published spending series shows a cut. Emarketer's March 2025 forecast put US influencer spending on track to pass $10 billion in 2025, landing at $10.52 billion. Its estimate for 2024 growth rose to 23.7%, an upgrade from an earlier 16.0% call. The firm's analyst Jasmine Enberg described the category as maturing and diversifying, not shrinking.
A trend about buying less did not stop brands from buying more creator time. The spending growth ran through the trend's whole first year. Budgets moved toward paid amplification of creator posts and toward channels outside social feeds. Nothing in the forecast treats deinfluencing as a drag on the line item. This connects to our earlier piece, Social Commerce Features: What the Evidence Actually Shows Three Years In.
What Controlled Brand Tests Found
Swayable, a creative-testing vendor, ran randomized tests on Stanley and Dyson deinfluencing TikToks in 2024. Its most impactful videos cut brand favorability by about 20 points and purchase intent by 17 points. Videos that pitched a rival product did less damage, in the range of 3 to 6 points. The firm said no other asset among more than 1,500 it tested that year moved opinion that far.
Complaints about product quality and customer service drove the worst outcomes. Respondents in the test recalled a failing attachment and an ignored support ticket more than any pitch. A claim backed by a visible defect beat a generic warning every time.
Where The Hype Outran The Numbers
Engagement never predicted damage. Swayable found that view and like counts did not track persuasive force. Every video among its top three performers had fewer than 2,000 likes. The most popular clip in the test cut favorability by 8 points, which ranked fifth. A team that screens creators by reach alone would have missed each high-impact threat.
The vendor caveat applies here too. Swayable sells pre-testing software, so findings that make testing urgent flatter its own toolbox. Its sample covered two consumer brands in viral product categories. Treat the point estimates as one lab's results, not a census of TikTok.
Conclusion: What Two Years Settle
Three things survive the two-year check:
- Creator spend grew through the trend's run: a 23.7% rise in 2024 and a $10.52 billion mark for 2025, per Emarketer.
- The brand risk is real but conditional: the strongest test clips moved favorability by about 20 points, while rival-pitch clips moved it 3 to 6 points.
- Reach is the wrong alarm: the most damaging clips in the test had under 2,000 likes.
The honest read is that deinfluencing became a testing discipline, not a spending event. Brands that took the hit now vet creative with panels and controlled exposure rather than follower counts. That habit will outlast the trend that forced it.
