A travel and food editorial publisher in Indonesia and Thailand activates a single mid-article outstream unit alongside existing display monetization. Over six weeks, article-level RPM lifts 58% with no measurable bounce-rate change and slight improvements to Core Web Vitals.
The publisher operates a travel and food editorial site serving readers across Indonesia and Thailand. Editorial style is long-form — feature articles run 1,800–2,400 words with substantial photography. Reader dwell time averaged 4 minutes 12 seconds at the engagement start; bounce rate was 38%, which is healthy for editorial content of this type.
Monthly pageviews were approximately 3.2 million with a 78/22 mobile/desktop split. Monetization at the engagement start was display-only: one 300×250 in-content rectangle and one 300×600 sidebar unit. Average article-level RPM was $1.94. The publisher had not previously activated video monetization because the editorial team has no video production capacity.
The conversation about outstream began when the publisher noticed competitor sites running outstream units in similar editorial layouts. The publisher's question was direct: can we earn video CPMs without producing video content? The answer is what outstream is built for.
Placement decisions for outstream are the difference between a unit that earns and a unit that drags. We considered three placement options: above-the-fold (after the hero image, before the first paragraph), mid-article (after the third or fourth paragraph), and below-content (after the article body, before related content).
Above-the-fold placement was rejected because the publisher's articles open with high-quality hero photography that drives reader engagement; replacing or competing with the hero image would have hurt session metrics. Below-content placement was rejected because the publisher's analytics showed that reader scroll-depth-to-bottom was 31% — too low to monetize a below-content unit effectively.
Mid-article placement was selected because the editorial structure is consistent (intro paragraph, body content with subheadings, conclusion). A unit placed after the third paragraph reliably falls inside the reader's attention window without breaking the editorial flow. The single unit was placed via a data-pc-slot div in the editorial template; the editorial team did not need to modify individual articles.
Week 1: outstream fill rate was 71% with average eCPM of $4.20. The combined article RPM (existing display + new outstream) lifted 22% above the previous baseline.
Week 2: fill rate 84%, eCPM $5.10. Article RPM 38% above baseline. The unit had developed a stable demand profile; the two highest-paying video demand sources (regional CPG and travel brand campaigns) were consistently competing for the inventory.
Weeks 3–6: article RPM stabilized at 56–61% above baseline. We activated the consent-gated personalized-demand path which lifted RPM on Singapore-traffic articles by an additional 8%. The final six-week average RPM lift was 58%.
Bounce rate during this period varied between 36% and 39%, which is within the publisher's normal week-over-week range and statistically indistinguishable from the pre-launch baseline. Dwell time was unchanged. Pages per session was unchanged. Core Web Vitals improved slightly: LCP improved from 2.41s to 2.28s because the outstream unit lazy-loads while the previous in-content display was eagerly loaded.
The case study illustrates an operating principle Pub Clarity holds across the platform: fewer, better-placed slots typically out-earn more, weaker-placed slots. The publisher could have added two or three outstream units per article; we recommended one. The single-unit approach captured the majority of the available yield while preserving the editorial reading experience.
A second illustration: contextual placement matters more than format choice. Outstream in the wrong placement would have produced a smaller RPM lift; outstream in the right placement produced 58%. The mid-article placement worked because the publisher's editorial structure created a reliable reader-attention moment for the unit to enter.
A third: conservative defaults compound. We activated the unit with muted autoplay, viewability-gated playback, and no-close. We could have activated with a more aggressive configuration (volume-on autoplay, frequency increased to two units per article, close button enabled). The conservative defaults earned 58% more than baseline without measurable user-experience cost. Aggressive defaults might have earned slightly more in the short term and risked the publisher's reader relationship in the long term.
Apply as a Pub Clarity publisher. After 90 days we offer to write up your monetization journey — anonymised or attributed, your call.
Apply as a publisher →