What the data can, and cannot, show.
The findings rest on public data alone. This page sets out what was collected, how it was validated, and the metrics that public access cannot reach, so every conclusion can be weighed against its evidence.
Accounts and handles in scope
The ten brands and the exact public handles analysed on each platform, as provided at the start of the engagement. Perfetti Van Melle's two brands are tagged in blue.
| Brand | TikTok | YouTube | Group | |
|---|---|---|---|---|
| Chupa Chups | @chupachupsarabia | @chupachupsmiddleeast | @Chupachupsarabia5513 | Perfetti Van Melle |
| Mentos | @mentosmiddleeast | @mentosmiddleeast | @MentosArabia | Perfetti Van Melle |
| Bazooka | @bazookaarabia | @bazooka.arabia | @BazookaArabia | Competitor |
| Borgat | @myborgat | @myborgat | @myborgat8621 | Competitor |
| Extra | @extra_arabia | @extraarabia | @ExtraGumArabia | Competitor |
| Haribo | @haribohappyworld_uae | @haribo.arabia | @hariboarabia | Competitor |
| Juicy Drop | @juicydroparabia | @juicydroparabia_ | @JuicyDropArabia | Competitor |
| Nova | @thenovagum | @thenovagum | @novagum3114 | Competitor |
| Skittles | @skittles | @skittles | @SkittlesArabia | Competitor |
| Zello | @myzelloworld | — | — | Competitor |
A dash means no public account on that platform. Snapchat handles were in scope but the connector was not in place this round, so Snapchat is not covered here. Skittles is the global @skittles account, with SkittlesArabia on YouTube the only regional handle provided.
What is public, and what is not
| Metric | Public? | How it is treated |
|---|---|---|
| Follower / subscriber count | Yes, snapshot | Current count per account |
| Net follower growth, churn | No public history | Posting cadence over time used as a proxy |
| Reach, impressions, reach rate | Owner only | Not shown |
| Video views | Partial | Public on TikTok, YouTube and Instagram Reels |
| Likes, comments, saves, shares | Yes (saves and shares on TikTok) | Full engagement analysis |
| Organic vs paid split | No | TikTok promoted flag and the Meta Ad Library used as signals |
| Spend, CPM, CPC, ROAS, CTR, conversions | Owner only | Not shown, never estimated |
| Audience demographics | No | Read directionally from comment language and dialect |
| Snapchat | Not connected this round | Outside current coverage: the data connector was not in place, so Snapchat is not assessed here and can be added in a later pass |
Five steps
Collection
Public posts across Instagram, TikTok and YouTube for every active account over twelve months, plus every public comment, 18,397 in total. Snapchat sits outside this round, as the connector for it was not in place.
Completeness
Some accounts returned partial data on the first pass. Each was collected several times, merged and de-duplicated by post ID, then checked against its reported post count.
Attribution
Instagram posts are credited to the account they appear on, not the technical author, so co-authored creator posts count as the brand's content and are tagged solo or collab.
Engagement
Reported two ways and never blended: engagement by followers for comparison, and engagement by views as the truer quality measure on video.
Comment analysis
Language, dialect, sentiment and recurring themes were classified across all 18,397 comments. Language split is reliable, Gulf-versus-other dialect moderately so, and the country split is directional and labelled as such.