Seeding video views: if a video is low on day 3, will it take off if you wait?

We followed the views of 2,493 videos from 14 seeding campaigns run by PLAYLAB for 30 days. Of the videos in a campaign's bottom 25% on day 3, 83% were still in the bottom 25% on day 30. Of the 2,235 videos under 1,000 views on day 3, five (0.2%) passed 10,000 views within 30 days. A single video is largely settled by day 3, but a campaign's total views on day 3 were only 64% (median) of its 30-day total.

About three days after seeded videos start going up, you are looking at a views report. Several videos have stalled at a few hundred views. The question is whether to write them off, or to wait in case the TikTok algorithm picks them up later.

PLAYLAB records the views of campaign videos every day. To answer the question, we took the 2,493 videos posted between June 18 and September 4 that have both a day-3 and a day-30 record.

For a single video, most of the result is in by day 3

Looked at one video at a time, views did not grow much over 30 days.

• Median views were 308 on day 3 and 439 on day 30.
• The median video's day-30 views were 1.35x its day-3 views. 15% of videos more than doubled, and 1% (26 videos) grew more than tenfold.
• For the median video, 74% of the views it earned over 30 days had already come in by day 3.

Rankings barely moved either. Within each campaign, we split videos into four groups by day-3 views, then split them again on day 30.

Day-3 position (within the campaign)	Same group on day 30	Moved up to the top 25%
Bottom 25% (630 videos)	83%	3.3%
Top 25%	80%	—

Eight in ten videos in the bottom group on day 3 were still in the bottom group a month later. Looking at each of the 12 campaigns with 30 or more videos, this share ranged from 76% to 92% — the same pattern in every campaign. About one in thirty bottom-group videos reached the top 25%. If the question is whether one particular video will take off if you wait, PLAYLAB's data mostly says no.

Late breakouts are rare, but large when they happen

2,235 videos had fewer than 1,000 views on day 3. Within 30 days, five of them (0.2%) passed 10,000 views and 13 (0.6%) passed 5,000. Narrowing to the 1,203 videos under 300 views on day 3, 34 (2.8%) passed 1,000 views within 30 days.

Working backwards from the 48 videos that had more than 10,000 views on day 30, 43 had already passed 1,000 views on day 3. Most of the videos that ended up big were already visible on day 3.

The rare exceptions were large. The video that grew most went from 270 views on day 3 to 860,000 on day 30. That one video accounted for 38% of all the views the 2,493 videos gained after day 3. Nothing in the day-3 numbers said in advance which video it would be.

A campaign total is only two-thirds in on day 3

A single video is mostly decided within three days, but a campaign's total views behave differently. The top videos keep growing, and the occasional late breakout pushes the total up.

Across campaigns, the median day-3 total was 64% of the 30-day total, ranging from 28% to 74% depending on the campaign. By day 7 the median was 77% (65–87%). Reporting a campaign on day-3 numbers can undercount the 30-day result by roughly a third.

Splitting the decision in two makes it less likely to be wrong

These results do not give one answer to "can you judge on day 3?" The answer depends on what you are judging.

Picking individual videos on day 3 is usually fine. When choosing videos for paid ads or creators to work with again, 80% of the day-3 top videos were still in the top group a month later. Rankings rarely flip, so choosing early costs little.

Judging the whole campaign gets more accurate the longer you wait. Only about two-thirds of the total is in on day 3, and the gap comes from a handful of videos nobody can predict. State in the report how many days after upload the numbers were taken, and re-count once in weeks two to four to catch the videos that grew late.

What these numbers do not tell you

• This analysis cannot tell why a late breakout broke out. The brand or the creator may have run ads separately. We excluded videos with a Spark Ads code registered with PLAYLAB, but could not check for other ad spend.
• Videos deleted or made private within 30 days have no day-30 record and are not included. This can make growth look slightly larger than it really is.
• These are mostly TikTok videos by creators outside Korea, posted between June and September. Other platforms or periods may differ.

Data basis

• Scope: TikTok videos posted between June 18 and September 4, 2026 in 14 seeding campaigns run by PLAYLAB, limited to the 2,493 with both a day-3 and a day-30 views record. Videos excluded from their campaign and videos with a registered Spark Ads code were removed.
• Day-3 and day-30 views: the last views count recorded between 1.5 and 3 days, and between 25 and 30 days, after upload.
• Bottom and top 25%: position within the same campaign, ranked by views and split into four groups.
• Campaign total share: for the 12 campaigns with 30 or more videos, the sum of day-3 (or day-7) views ÷ the sum of day-30 views.