Spotify
Figure 1. Spotify’s Discover Weekly playlist personalized for useres
Spotify’s recommendation engine includes playlists like Discover Weekly (Figure 1), Daily Mixes, On Repeat, which optimizes for listening time, directly driving subscription retention—the single most important financial lever. A small percent improvement in retention can translate to millions in ARR. As such, the ROI from personalized recommendations is extremely high. Every incremental hour listened reduces churn risk and raises LTV. Personalized playlists are cheap to scale compared to the revenue uplift they generate.
Figure 2. LinkedIn’s personalized job recommendations
LinkedIn’s ranking models personalize users’ feed, job recommendations (Figure 2), and professional graph. Their north star metric is session frequency since more frequent returns mean more ad impressions and higher conversion to LinkedIn Premium plans. As such, the ROI from personalization is also pretty strong. Better feed relevance increases DAU/WAU, which increases the surface area for ads, messages, and recruiter interactions—driving ad revenue, B2C plan upgrades, and B2B subscriptions.
Tiktok
Figure 3. Tiktok’s “Why this video?” feature shows users why this video was recommended to the user
TikTok’s For You Page learns your preferences based on videos you watch within seconds (Figure 3). It optimizes for watch time per session, which drives the ad auction and boosts creator distribution. The more precise the personalization, the more valuable each ad impression becomes. Tiktok’s ROI from personalization is industry-leading compared to other social media platforms. TikTok’s CPMs remain high because their targeting is behavior-based, deeply predictive, and most importantly, real-time. Even small gains in watch time meaningfully increase total ad inventory and auction efficiency.
