Volume held up on TikTok, but the typical post landed weaker on both platforms

Coverage

What was read, and where reading stopped

Both accounts were read newest-first from their public post lists. Reading continued past the start of the comparison period on both platforms, so every September post is included and no figure depends on a partially covered month.

Instagram reading covered five pages of 12 posts each; TikTok covered five pages of 15 videos (three videos repeated across page boundaries and were counted once). Every September figure therefore rests on a complete month on both platforms.

Profile details as read

In one paragraph

The answer a client can read on its own

Overview

Likes fell faster than comments on Instagram; TikTok's totals collapse is one video, not a trend

Figures below are per platform and are deliberately limited to posts, likes and comments. Views and engagement rates are not combined across platforms, because Instagram and TikTok count them differently and a single total would not mean anything.

Two single posts drive the totals more than the median does. On Instagram one 10 September Reel accounts for 44.4% of September's total likes (54,782 of 123,513); on TikTok one 12 September video accounts for 96.8% of September's total views (6,300,000 of 6,505,694). Medians per post are the more honest comparison, and both are reported below with the number of posts they rest on. The 62.3% rise in Instagram's median comments is inflated by giveaway posts, flagged in the ranking section.

Scorecards

Each platform's own numbers, with the base behind every median

Every figure below is shown for October, for September and as a change. The base column gives the number of posts each median rests on. All medians in this report rest on at least 5 posts, so none is marked too few to judge.

Best & weakest

Three strongest and three weakest posts per platform — with what the captions and formats show

Instagram is ranked by engagement as (likes + comments) ÷ followers; TikTok is ranked by views, with each video's engagement shown beside it. Flagged posts stay in the totals but are not used to draw conclusions about which format, topic or platform works.

Instagram · strongest three

Instagram · weakest three

TikTok · strongest three by views

TikTok · weakest three by views

Flags applied across the reporting period

• Giveaways on Instagram (3 of 14 posts): 12 October Momofuku (746 comments), 9 October Lonely Ghost closed giveaway (1,161 comments) and 5 October sneaker giveaway (457 comments). Together they supplied 2,364 of the month's 5,083 comments — 46.5%. They inflate the median-comment and engagement figures, and none is used as evidence about format or topic.
• Paid partnerships: every October Instagram post carrying a paid-partnership or #ad-type label was owned by another account and is excluded from this report, so no retained Instagram post is a labelled ad. No retained TikTok caption carries an ad or sponsored label; six name partner brands (@uno, @LONELY GHOST, @Califia Farms and @Torani, @beauthistle, @Compartes), which is a collaboration mention rather than a declared paid partnership.
• Outliers that may have had paid promotion: the 1 October Instagram photo (15,086 likes against a 1,998.5 median) and the 6 October Instagram photo (8,097 likes) both show likes far out of line with their comments. Public numbers cannot show whether promotion was bought, so these two are flagged and excluded from format and topic conclusions.

What was posted

Every post in the reporting period, with its figures and link

One row per post, newest first. A dash means the platform does not show that figure for that post — Instagram exposes shares for Reels only, and views only for video. Flags are explained at the end of the previous section.

Instagram · 14 posts

Carousel and photo rows show no shares because Instagram reports share counts for video posts in this data; each carousel's caption and totals are shown, not its individual slides.

TikTok · 29 videos

Saves are TikTok's own save count. Likes, comments, shares and saves are all public counts as displayed at collection time; none of them is reach.

Two platforms

TikTok carried the volume; Instagram's typical post slipped further from its own baseline

Each platform is measured against its own September, not against the other. Raw likes and views are never ranked across the two, because a view on TikTok and a like on Instagram measure different things and the follower bases differ.

How the typical post on each did against its own September

TikTok's typical engagement held up far better than its typical view count: the median video drew 37.7% fewer views but converted almost as well as in September, down 6.6%. Instagram lost ground on both counts, with median engagement down 26.6% and median likes down 34.8%. Read that as each account drifting from its own baseline, not as one platform outperforming the other.

Topics that appeared on both platforms

A third shared theme was Caramel Apple flavour content: Instagram's 16 October recipe Reel drew 1,815 likes and 69,811 views, while TikTok's 20 October Crisp-versus-Caramel video drew 2,442 views and 115 likes. Pumpkin-season content also ran on both, with Instagram's 30 October carousel at 3,618 likes and TikTok's 23 October recipe at 3,343 views.

Next month

Three actions, each with the number to check in November

Each action is tied to a figure from the account's ordinary posts, not to a flagged post, and names the measure to read in next month's report. No outcome is forecast.

1. Instagram · lift carousels from 4 to at least 6 posts

   The three October carousels that were not giveaways had a median engagement of 0.81%, against 0.33% for Reels — about 2.4×. Carousels were also the only format whose count rose while total posts fell.

   Check next month: median engagement on carousels versus Reels, and whether carousels still hold above the 0.496% all-post median.

2. Instagram · publish three recipe Reels using the format that travelled

   The two October recipe Reels — 27 October at 68,880 views and 16 October at 69,811 views — both beat the Reel median of 40,143.5, which the month's Reels as a whole missed by half against September.

   Check next month: median views on recipe Reels against the October Reel median of 40,143.5.

3. TikTok · hold output at about 29 videos and stop adding volume

   Publishing rose 20.8% while median views fell 37.7% to 2,761, so extra videos did not lift the typical one. The strongest unflagged videos were short joke formats at 5–10 seconds (4,151 and 4,150 views), which is where a reallocated slot is best spent.

   Check next month: median views per video against 2,761, and whether it moves without another increase in volume.

Ready to send

Email you can paste to the client

Sources & limits

What this report rests on, and what it cannot show

Sources: instagram.com/drinkolipop and tiktok.com/@drinkolipop, public profile and post data retrieved through Nolyn on 9 November 2026 between 16:18 and 16:23 JST (07:18–07:23 UTC). All per-post figures, totals, medians and percentage changes were computed from the retrieved per-post values.

• Follower counts are current, not historical. Engagement in both periods is divided by the follower count read on 9 November 2026 (478,915 on Instagram, 589,913 on TikTok), because past follower counts are not public. Follower growth is therefore not reported.
• Late-October posts have had less time to collect. Posts from 25–31 October had roughly one to two weeks to accumulate likes and views when the data was collected, against a full month for early-October and September posts, so the end of the reporting period is under-recorded relative to its start.
• Not public and not included: reach, impressions, Instagram saves, profile visits, Stories, link clicks and ad spend. Nothing in this report is reach; likes and views are counted as likes and views.
• Exclusions: 21 Instagram collaboration posts owned by other accounts were left out, along with 0 posts with hidden like counts; 0 TikTok videos were left out. Their engagement sits in other accounts' figures, not OLIPOP's.
• Outlier distortion: September's totals are dominated by one Instagram Reel (44.4% of the month's likes) and one TikTok video (96.8% of the month's views), so totals understate how the ordinary post did. Medians are used for that reason.
• Possible paid promotion cannot be confirmed. Two Instagram posts show likes far out of line with their comments and with the account's usual range. Public numbers cannot show whether promotion was bought, so they are flagged and excluded from conclusions.
• Carousel contents: the data gives each carousel's caption and total counts, not its individual slides, so no claim is made about the performance of a particular slide.
• Two platforms, not one. Because both accounts were covered, this report compares them only on publishing volume and on each platform's own typical-post trend — never on raw likes or views.