Story of the month
Views +100% · Follows +178%
Page CSV shows total views rising 100.1%, from 728,238 to 1,457,564, while net follows increased from +558 to +1,554. In Content CSV, however, the top two posts account for 46.7% of views and 85.2% of net follows. Removing the top post within the same Content dataset leaves 798,626 views, up 20.4% from July, but median views fell 28.0% from 6,775 to 4,878, interactions fell 7.2% to 15,020, and net follows fell 26.7% to +354. Total reach expanded, while efficiency and follower acquisition across regular posts weakened.
Performance by Post Type
All figures from Content CSV. Because Page CSV covers different activity, the total row is a Content total, not the monthly total
| Type | Posts | Total Views | Avg Views | Median | ER | Net Follows | Revenue | RPM |
|---|
| Reels | 19 | 753,928 | 39,680 | 7,295 | 1.93% | +1272 | $34.50 | $0.0458 |
| Photo | 59 | 476,801 | 8,081 | 4,593 | 1.24% | +83 | $47.42 | $0.0995 |
| Text | 1 | 1,082 | 1,082 | 1,082 | 1.11% | +0 | $0.06 | $0.0555 |
| Content total | 79 | 1,231,811 | 15,593 | 4,900 | 1.66% | +1355 | $81.98 | $0.0666 |
How to read this
The two-layer structure seen in July — video wins follows, photos win revenue efficiency — still holds. But average views for video are pulled hard by the top two posts, so type-level averages should not be read as the level of a normal post. That is why medians are shown alongside.
Top Posts (Top 5 by views)
The top post accounts for 35.2% of Content views, and the top two for 46.7%. The leading post covers foreign-resident regulations, but its link to the Japan-based audience remains a hypothesis
1
How will Japan tighten regulations on foreign residents?
08/01/2026 · 23s · Reels
Share of Content views
35.2%
2
PM Takaichi responds to U.S. sanctions on the ICC chief
08/19/2026 · 11s · Reels
Share of Content views
11.6%
3
Typhoon No. 13 forecast to affect Ogasawara and Okinawa
08/01/2026 · Photo
Share of Content views
5.4%
4
Healthcare heroes stay with patients during an earthquake
08/07/2026 · 91s · Reels
Share of Content views
5.3%
5
U.S. involvement in yen-buying intervention revealed
08/01/2026 · Photo
Share of Content views
2.6%
Post themes are summaries of the Khmer originals, not literal translations.
Key Observations
Excluding the outlier changes the conclusion metric by metric
A Page CSV total must not be adjusted by subtracting a post-level value from Content CSV because the datasets cover different activity. Removing the top post within Content CSV leaves 78 posts, 798,626 views, 15,020 interactions, +354 net follows, $58.65 in revenue, and a median of 4,878 views. Compared with all July Content posts, views are up 20.4% and revenue is up 9.6%, while interactions are down 7.2%, net follows are down 26.7%, and median views are down 28.0%. The result is not simply flat; reach volume and regular-post efficiency moved in different directions.
Photos added total volume, but marginal efficiency declined
Photo output rose 84%, from 32 to 59 posts. Total photo views increased 39%, from 341,887 to 476,801; interactions rose 43%, from 4,122 to 5,895; and revenue rose 30%, from $36.43 to $47.42. It is therefore incorrect to say the added photos had no effect. However, average views fell 24%, from 10,684 to 8,081, median views fell 28%, from 6,358 to 4,593, and net follows rose only 8%, from +77 to +83. Higher output added reach, interactions, and revenue, but marginal efficiency per post declined.
Broader distribution alone does not explain the ER decline
Page ER fell from 3.00% to 1.84%. The rise in Reels traffic to 86.2% is consistent with the hypothesis that broader non-follower distribution diluted response rates, but July's 62.1% figure covers a 28-day view while August covers a calendar month, so the comparison is not exact. Within Content CSV, ER after removing the top post is 1.88%, and Reels ER after removing that post is 2.84%, below July's 3.87%. Broader distribution is therefore a plausible contributor, but weaker response efficiency across regular posts cannot be ruled out.
The breakout post monetized at a lower rate, cause unknown
Page RPM fell 21.6%, from $0.0804 to $0.0630. Within Content CSV, the top post generated an RPM of $0.0539, compared with $0.0734 for all other August posts combined, confirming that the breakout post monetized at a lower rate than the rest of the month's content. The cause cannot be isolated because post-level viewer geography, follower status, ad impressions, and eligible inventory are unavailable. Viewer location, short duration, placement, and ad inventory should remain hypotheses rather than conclusions.
The short-versus-long pattern stays descriptive
Within August, shorter Reels reached more with lower ER and longer Reels reached less with higher ER. But the ≤15s band holds one post, 16–30s holds two, and 31–60s holds one, so length cannot be separated from topic, posting time, or news intensity. Only the 61–120s band (12 posts) has a usable sample, and it represents normal operating level. Reassess once several short Reels under comparable topics and conditions have accumulated.
The link between Japan-resident share and the top post remains a hypothesis
The lifetime share of followers based in Japan rose from 21.0% to 24.6%, while the top post covered tighter regulations affecting foreign residents. This is consistent with the hypothesis that practical policy information reached Cambodians living in Japan, but post-level viewer geography is unavailable, so neither causality nor the audience mix can be confirmed. Reassess reproducibility once more posts on comparable themes have accumulated.
Data Quality and Interpretation Limits
The Page-to-Content gap widened, and its breakdown is unknown
August Page CSV contains 1,457,564 views, while Content CSV contains 1,231,811, leaving a gap of 225,753 views, or 15.5% of the Page total. This is larger than July's gap of 64,705 views (8.9%). June, however, showed 61,139 (13.4%), so the gap fluctuates by month rather than growing steadily. Possible sources include activity on older posts, distribution surfaces not represented in the Content export, publication-date boundaries, and extraction timing, but the breakdown is not known. Page CSV should therefore be used for monthly totals and Content CSV for post-level comparisons; values from the two sources must not be subtracted from one another.
Two boundary records sit in the August Content export
August Content CSV contains two boundary records whose displayed publication date is July 31. A 76-second Reel at 21:38 contributes 8,860 views, 453 interactions, +16 net follows, and $0.83 in revenue. A Photo at 23:51 contributes 2,652 views, 41 interactions, no net follows, and $0.33 in revenue. Combined, they account for 11,512 views, 494 interactions, +16 net follows, and $1.16. Facebook may use a different timezone for period filtering than for the displayed CSV timestamp, but this has not been confirmed. The 79-post count should be treated as the records included in Facebook's August export, and neither record should be removed solely because of its displayed date.
Two sources report different net-follow totals
Page CSV sums to +1,554 while the Audience screen displays +1,569, a difference of 15. This is presumably a definitional difference in aggregation, but it has not been confirmed. This report uses the CSV figure throughout.