Buddy Bites HK · Instagram performance review
Two analyses answering one question from opposite ends. Part I reads the fifty posts Buddy Bites published itself. Part II reads the 212 posts other accounts published about Buddy Bites. Every post in both halves is scored against what it would normally have achieved anyway — its own format, its own channel, its own account's baseline — so nothing looks strong just because it was handed the biggest audience.
Every figure in this report is calculated from the live content tracker — four tabs covering own-account posts, collaborations, tagged posts and collaborator profiles: Buddy Bites — Own Account Content Tracker ↗
Fifty posts Buddy Bites published itself — 27 on its own feed and 23 native collaborations — each scored against what its own format and channel normally deliver, so a pillar can't look strong just because it always got the Reel slot.
Read these first. Everything after this section explains where they come from and what to do about them.
Buddy Bites has two different engines on one Instagram account. The own-account feed is a steady, low-amplitude brand layer. The collaboration programme is the actual growth engine — and inside it, a single content pillar is doing almost all the work.
Collaborations are 46% of posts and 96% of all engagement. That ratio is distorted by one enormous post — so the fairer test is the median, which is immune to outliers. It tells the same story with less drama.
One post carries 74.8% of all collaboration engagement, and the top three carry 91.7%. That is a textbook power law, and it means average-based reporting on this programme is meaningless.
But strip the giant out and the median barely moves — from 141 to 138. So the typical collaboration still beats the typical own-account post 3.4×. The programme is not one lucky hit propping up a dud strategy; it is a reliably better channel that also happens to produce occasional monsters.
Seven content pillars, defined by subject rather than format: Community Care / Social Good (shelter donations, rescue partnerships), Brand Love & Engagement (behind-the-scenes, holidays, brand-identity posts with no purchase angle), Trust & Proof (testimonials and third-party validation), Product/Purchase Education (product info, discounts, purchase funnels), Nutrition & Safe Ingredients (ingredient and nutrition explainers), Taste & Pet Enjoyment (pets tasting or enjoying the food), and Convenience for Pet Parents (ease-of-use and delivery). Each post is read individually and coded to whichever pillar its caption is actually about — a post that genuinely covers two is tagged with both and counts fully toward each.
Each post is scored as an index against the median of its own format-and-channel cohort: 1.00 means "exactly what that slot normally delivers," 2.00 means double. Plotting that against how much of the calendar a pillar occupies turns performance into a decision.
† Baseline caveat: Trust & Proof and Taste & Pet Enjoyment (marked † above) live almost entirely in Collaboration — the same slot Community Care dominates. Because Community Care is exceptionally strong, it drags up the "typical" bar that everyone sharing its slot gets measured against. Recomputed against a baseline with Community Care posts removed, both jump to ~1.01× — essentially typical, not underperforming. Community Care is the real standout; Trust & Proof and Taste & Pet Enjoyment aren't weak, they're just benchmarked against it. Pillars living mainly in Own Account (Brand Love, Product/Education, Nutrition) don't carry this caveat — Community Care never shares their slot, so their index is read as-is.
Convenience for Pet Parents shows the second-highest index (1.86) — but on two posts. That is a hypothesis, not a finding. It belongs in the next test cycle, not in a strategy deck. Every row in the table carries its own confidence flag for exactly this reason.
Views and engagement rate answer different questions. Views measure how far Instagram pushed a Reel; engagement rate measures whether the people it reached actually cared. The two diverge sharply — and the gap is where the useful decisions live.
The loved but buried quadrant — high engagement rate, below-median views — is seven Reels that people demonstrably liked and the algorithm under-distributed. That is the cleanest paid-promotion shortlist in the dataset, and it costs nothing to produce.
The one confirmed boosted Reel in the dataset — DW8gG4Fyz-v, 24,797 views at 0.20% — lands squarely in wide but flat, with a performance index of 0.96. Paid money bought it reach; engagement stayed at exactly baseline. That is the signature of paid distribution.
Content performance by month, split by channel. Not account growth — follower history isn't available — so this reads as "is our content getting better," not "are we getting bigger."
Collaboration Reels climb from 1.96% in January to 3.19% in July — the programme is getting better at picking and briefing partners. Own-account Reels sit flat between 0.70% and 1.14% for seven straight months. The owned feed isn't deteriorating; it just isn't learning.
What this analysis can and cannot support. Stated up front, because a strategy recommendation is only as good as the reader's ability to check it.
212 posts from the 158 accounts that tagged Buddy Bites — creators, venues, rescue organisations and customers — each read against what that account normally achieves, so an 810-follower customer and a 500,000-follower celebrity can be judged on the same scale.
Two measures carry this report, and they answer different questions. Keeping them apart is what makes the rest of the analysis readable.
One limitation, stated up front. A meaningful share of pet-focused creators and other creators hide their like count on Instagram. Public data then records a near-zero engagement total instead of the real, unknown one. Every figure for those two categories in this report is therefore a floor rather than an exact count. There is no way to recover the true number from public data — only the creator can supply it. Venue / Business, NGO / Rescue Org and Customer Advocate are unaffected.
Where the posting effort goes, against where the response comes back.
Share of posts
212 posts by partner type
Share of engagement
79,034 engagements by partner type
Reading the two together
Return ratio = share of engagement ÷ share of posts. Above 1.0 means the type delivers more response than its share of the posting effort.
| Partner type | % of posts | % of engagement | Return ratio | Ratio excl. @wadahiromi |
|---|
Caveat. Pet-focused and Other Creator figures in the chart and table above are a floor, not a precise count — a share of those posts come from creators who hide their like count, which reads in public data as near-zero engagement. Both categories' true share of engagement, and their return ratios, are higher than shown, likely by a real margin. The comparison stands as the best available read rather than an exact one.
Scale sizes a campaign. Affinity judges the creative.
Scale — median engagements per post
Account-level medians
Affinity — against the account's own norm
Dashed line marks parity at 1.0
Content depth — how much of the post is actually about Buddy Bites
Share of each partner type's posts, from 194 tagged posts scored by reading every caption individually. Hover a bar for the full four-level breakdown.
Set this against the Affinity chart above and the two rankings barely differ. The three types with the highest Affinity — Venue / Business, Customer Advocate, NGO / Rescue — are the same three, in the same order, whose posts least often make Buddy Bites the subject. Some of that is simply how audiences behave — people engage less with anything that reads as promotion, so a measure like Affinity will always give some credit for saying less. Read the other way round, though, it points at an opportunity rather than a problem: the posts built entirely around Buddy Bites are exactly where stronger creative has the most to gain. It is also why this report compares within a partner type rather than across them.
Caveat. Pet-focused and Other Creator Scale figures understate the true median, for the same hidden-like-count reason described in Section 1. Their position on the Scale chart should be read as a lower bound.
The accounts that delivered, the posts worth studying, and the pattern underneath them.
This section moves from categories to individual posts. It identifies which accounts produced the most engagement, which specific posts beat their own account's norm, and — from reading those posts individually — what the strong ones have in common.
The eighteen accounts that produced the most publicly recorded engagement, with each account's own baseline engagement rate alongside the rate its Buddy Bites posts achieved. @wadahiromi is a genuine outlier — a 502,000-follower celebrity, far larger than anything else Buddy Bites works with — and is flagged rather than treated as representative; she is excluded from the group medians used elsewhere in this report. Accounts whose only Buddy Bites post is that same shared collaboration are not listed, so every row here reflects content the account produced itself.
| Account | Partner type | Followers | Account ER | BB post ER | Affinity | Posts | Engagements | Links |
|---|
Account ER is that account's normal post engagement ÷ followers. BB post ER is the same calculation on their Buddy Bites posts — marked ▼ where it sits below the account's own rate, meaning Buddy Bites content underperformed what that audience usually engages with. Highlighted rows have Affinity above 1: the post beat that account's own norm, and the pairing is worth repeating.
Filtered to posts where Buddy Bites genuinely occupies the post, tagged to a single partner, on accounts with a stable baseline. Venue / Business and Customer Advocate are excluded here — the first because its posts aren't about the product, the second because those are organic customer posts rather than partnerships. 111 posts qualify; the top five in each category by Affinity are shown. Posts marked BOTH also rank in that category's top five by raw engagement — they beat their own account's norm and delivered volume, which makes them the strongest models to brief against.
Every post above was read individually. Two patterns emerged — and they are genuinely two patterns, not one formula applied twice.
1 · Rescue organisations and customer advocates — heartfelt storytelling wins, told in words or not
Every one of NGO / Rescue's strongest posts centres a specific, individual story. @straypawsalliance's top post is a rescue dog beside a wall of donated food — no caption needed to feel the contrast. @tails.lantau's two strongest posts are both framed as an arc: a dog Buddy Bites helped raise, that then found a home. The category's weakest posts, by comparison, are generic thank-you announcements with no story attached — same accounts, same cause; the only real variable is whether there is a specific animal and a beginning-to-end, or a press-release-style acknowledgement. The clearest single data point is @tmahk, a customer with 810 followers, whose real adoption story reached 31.03 Affinity — the highest figure in the entire dataset, ahead of every influencer and the celebrity post.
2 · Pet-focused creators — no universal formula; it is the account's own character plus execution
Pet-focused accounts do not lean on storytelling the way rescue organisations do. Each has already built its own character with its audience — a look, an aesthetic, a pet's own personality — and the posts that outperform are the ones that attach Buddy Bites to that existing identity rather than working against it: a cat popping out of the product box, a dog visibly reacting to the food, packaging that matches the account's palette, the pet doing the trick it is already known for. There is no single content formula to copy here. The work is to identify what already catches that specific account's audience, then find a way to attach Buddy Bites to it. For Reels, the opening few seconds carry the post — every top performer in this category hooks the viewer immediately rather than building up to it.
Other Creator is a partial exception to both patterns. Outside @wadahiromi's celebrity collaboration, the strongest posts here had no identifiable content driver on review — the lift appears to come from the account's own audience trust rather than anything specific in the post. That said, this reading rests on a small pool of posts, too small to draw a firm conclusion from — treat it as a hypothesis to watch, not a rule to brief against yet.
Neither is proven. Both are inexpensive to test.
These sit outside the main analysis because each rests on a smaller slice of the data. They are included because both point at a decision that is currently being made by default rather than deliberately.
Native collaborations outperform plain tags
Median Affinity by creator size band
| Creator size | Collab | Tag |
|---|
Collaboration posts beat plain tags in every size band, most sharply at the two ends. Converting more tags into native collaborations is the obvious move — but it has to be weighed against feed consistency and audience fatigue, and against the fact that some accepted collaborations were later removed from the creator's feed.
The cat gap
Species focus of the roster, against a cat line launched in 2026
| Focus | Accounts | Posts |
|---|---|---|
| Dog | 72 | 88 |
| Cat | 19 | 21 |
| Both | 6 | 8 |
| Pet, unspecified | 23 | 57 |
| Non-pet | 38 | 50 |
Cat accounts receive under a quarter of the posts dog accounts do. The roster was built for the dog range and has not yet shifted to match the cat line — worth weighing against how much revenue that line is expected to carry.
What this analysis can and cannot support. Stated up front, because a strategy recommendation is only as good as the reader's ability to check it.
How the measures are built
What this analysis cannot tell you