What makes TikTok creators reply?

The short answer

1. Write in their language. Creators written to in Spanish replied twice as often as those written to in English, and agreed a deal five times as often.
2. Go smaller. Under 5,000 followers, 30% replied; over 100,000, 19%. Creators under 20,000 agreed deals twice as often.
3. Send the follow-up. 26% of all replies came only after it.
4. Look for brand mentions. Creators who already name brands in their bio or captions replied 34% of the time, against 24%.
5. Don't trust fit scores to predict replies. An AI score of how well each creator fits the brief had no link to who answered.

1. Language beats everything else

Every email went out in the language the creator posts in. Spanish- and Italian-speaking creators replied far more and, more importantly, closed far more: 11% of the Spanish-speaking creators and 13% of the Italian ones agreed a paid video, against 2% of the English-speaking ones. English-speaking creators in the US and UK likely get the most brand mail.

2. Smaller creators answer, and say yes

Reply rates drift down as follower counts rise: 30% under 5,000 followers, 19% above 100,000. Deals fall faster. Creators under 20,000 followers agreed 4.7% of the time, those above 2.2%.

3. One follow-up, and patience

Half of all replies arrived within a day of the first email. Then there's a second wave: 26% came more than a week later, almost all of them to the follow-up. In total 113 of the 440 replies (26%) answered the follow-up, not the first email. A second follow-up brought back one creator.

4. Creators who already name brands

Creators whose bio or captions already mention brands in the niche have done paid work and expect the email. They replied 34% of the time against 24%, and agreed deals a little more often (5.0% against 3.6%).

5. Fit scores don't predict replies

Every creator was scored 0–100 by AI on how well their content fits the brief. The score is useful for picking who to contact, but it says nothing about who will answer: replies are flat across score bands, and the score's AUC against replies is 0.505 (0.5 is a coin flip). An estimate of how well each creator's audience matches the target did no better (0.484). The top band did agree deals more often (14 of 204), too few to call it.

What we changed in Trovaso

Trovaso already writes every first message in the creator's language and sends the follow-up for you. From this data it also learns, per account, who tends to answer — market, size, comments, brand mentions — and can sort your list by “most likely to reply”. On this data that model reaches an AUC of 0.587 on the newest quarter of emails, held out: modest, and much better than the fit score.

Method and limits

• 1,753 TikTok creators emailed through Trovaso by one consumer health app between 1 September to 15 October 2026, one first email and up to two follow-ups each, all in the creator's language.
• Reply = the creator wrote back to the thread. Deal = the conversation was marked as agreed. Emails after 15 October are left out so every creator had at least nine days to answer.
• One brand, one niche and one platform. Language overlaps with market and with how well the app itself sells there, so read the language gap as “these markets answer”, not as proof that translation alone does it.
• Correlations, not experiments. Group sizes are shown under each bar; the smallest (Italian, 64 creators) is the least certain.
• AUC: the chance that a random creator who replied ranks above a random one who didn't. 0.5 is chance, 1 is perfect.
• Data: creator-reply-study-2026.csv, aggregated counts only, free to reuse with a link (CC BY 4.0). Cite as: Trovaso (2026), “What makes TikTok creators reply?”, trovaso.com/research/what-makes-creators-reply.