Instagram placement typically beats Facebook on CPM by 15-30 percent for DTC brands but loses on ROAS by a similar margin, which is why platform-optimised placement rarely matches the all-in-one defaults buyers reach for. The gap is intent, not cost. Instagram traffic skews toward discovery and casual scrolling. Facebook traffic skews toward intent-driven consumption. A click from each surface is not the same click. Advantage+ Placement bundles them anyway and relies on the auction to allocate, which usually works but leaves 10-20 percent of performance on the table when the brand has reason to split placements manually.
Instagram wins CPM, Facebook wins ROAS, and the gap is intent not cost. Advantage+ Placement works for most accounts but costs 10-20 percent of performance where the intent gap matters. Audit placement ROAS before accepting the auto-allocation default.
Why does Instagram typically produce lower CPMs than Facebook?
Instagram typically produces lower CPMs because the surface has more impression inventory per active user than Facebook. The feed is faster, Stories cycle more frequently, and Reels deliver near-infinite scroll. More inventory at a given spend produces more impressions, which arithmetically lowers CPM. Facebook feed inventory is more constrained — the user scrolls less, posts stay visible longer, and ad load is lower per session. The impression supply difference drives the CPM gap mechanically. It has nothing to do with user quality on either surface.
Why does Facebook win on ROAS despite higher CPMs?
Facebook wins on ROAS because the clicks the surface produces come from users in a different mental state. Facebook feed consumption trends older, slower-paced, and more deliberate — users scroll with attention and click with intent. Instagram consumption trends younger, faster, and more passive — users scroll casually, double-tap habitually, and click with curiosity. A $5 Facebook click often converts at 2-3 percent. A $3.50 Instagram click often converts at 1-1.5 percent. The ROAS math produces Facebook as the winner despite the higher CPM, which is the structural reason platform-level ROAS reports almost always favor Facebook even in brands that intuitively feel more Instagram-native. The same intent-gap distinction applies to Reels versus Feed within Instagram itself, where the surface decides the click quality independent of creative.
When does Instagram actually beat Facebook on ROAS?
Instagram beats Facebook on ROAS in four specific cases. First, when the target audience is under 30 and buys on the platform they discover on — Instagram native. Second, when the creative is short-form video optimised for Reels and the Facebook feed version loses significant polish. Third, when the product is visually driven (apparel, beauty, home decor) and the Instagram grid context amplifies desire more than the Facebook feed does. Fourth, when the brand runs influencer-adjacent organic content that primes buyers for the ad — Instagram carries the primed audience, Facebook does not. In other categories (B2B, considered high-AOV purchases, utilitarian products) Facebook almost always wins on ROAS.
Should you split placements manually or let Advantage+ allocate?
Advantage+ Placement works for most accounts because the auction system allocates toward the surface delivering the best conversion signal in real time. The split happens automatically. Manual placement split beats auto-allocation in three cases: when creative is dramatically better on one surface than the other (a Reels-first video barely fits Facebook feed), when the brand needs placement-level reporting granularity for strategic decisions (not just performance optimisation), or when the audience strongly skews one platform (over 70 percent of past buyers active on Instagram, for example). For most scaled DTC accounts, auto-allocation captures 80-90 percent of the available performance and the manual effort to split is not worth the 10-20 percent gain.
How do you audit which placement is actually driving your ROAS?
1) In Ads Manager, break the last 90 days of campaign performance down by Placement (not just Platform). 2) Measure CPM, CTR, CVR, and ROAS by Instagram Feed, Instagram Reels, Instagram Stories, Facebook Feed, Facebook Marketplace, and Audience Network separately. 3) Identify the single highest-ROAS placement for each campaign. 4) If one placement is contributing more than 50 percent of ROAS at under 30 percent of spend, consider splitting it into a dedicated manual ad set with placement-specific creative. 5) If placements are roughly balanced across the top three, keep Advantage+ Placement — auto-allocation is already doing the work. 6) Rerun quarterly because placement economics shift with platform changes.
What is the common mistake brands make with placement optimisation?
The common mistake is designing creative for Instagram Reels and expecting the same ad to perform on Facebook feed. The vertical 9:16 format gets letterboxed on Facebook feed, the first-second hook that works on fast-scroll Reels is wasted on slower Facebook consumption, and the brand pays the CPM premium on Facebook without getting the ROAS advantage because the creative is miscalibrated for the surface. The right practice is to produce placement-aware creative — a Reels cut and a Feed cut from the same raw footage, each optimised for its surface — or to split placements manually when one creative version dominates the budget.
Conclusion
Instagram and Facebook deliver different kinds of clicks at different costs, and lumping them together in Advantage+ Placement is right for most accounts but costs 10-20 percent of performance where the intent gap matters. The practical next step is a 14-day placement audit comparing ROAS by placement on existing campaigns, then deciding whether to split based on where the gap sits for the specific creative and audience. In Advize's experience, roughly three in ten DTC accounts have a placement imbalance worth splitting manually.