How Bach AI Breaks Down Placement Performance Across Feed, Reels, and Stories

info wittelsbach
4 days ago
4 min read

Meta's default 'Advantage+ Placements' shows you a single campaign ROAS number — say, 2.8x. Drill into placement breakdown and you'll find that Feed is doing 4.1x, Reels is doing 1.9x, and Stories is doing 0.6x. The blended number hid the reality. You scaled the campaign, fed more budget to the placement losing money, and watched ROAS slowly bleed.

Placement-level performance is the most under-reported dimension in Indian D2C accounts. Bach AI splits every campaign by placement automatically, surfaces the action per placement, and flags when blended ROAS is hiding a problem.

Why Placement Performance Varies So Much

Feed, Reels, and Stories are different formats, different user behaviors, and different auctions. A creative that works on Feed often fails on Reels because:

Feed users scroll slowly and dwell on visual hooks; static images and 1:1 video work well.
Reels users scroll fast and engage with motion; static images often get skipped under 1 second.
Stories users tap through quickly; you have 1.5-3 seconds to land a message.
Auction density differs — Reels CPM is often 20-40% lower but CTR can be wildly variable.

A campaign with 'Advantage+ Placements' on doesn't fix this. It tells Meta to allocate budget across placements based on cost — not necessarily based on profitable conversions.

How Bach AI Breaks Down Placement Performance

Three views, automatically refreshed:

1. ROAS by placement

Feed, Reels, Stories, Right Column, In-Stream, Marketplace, and Search Results — each gets its own ROAS, CPM, CTR, and CPA. You see the placement-level breakdown for every campaign without changing any settings.

2. Spend distribution

Where is Meta actually sending your money? Many accounts assume Advantage+ Placements distributes evenly. In practice, 60-80% of spend often goes to one or two placements — usually Reels and Feed.

3. Creative-by-placement performance

Some creatives perform 4x better on Feed than on Reels. Bach AI shows you which creative is dragging which placement, so you can fix the creative or restrict the placement.

The Per-Placement Recommendation Card

Open any campaign and you see one card per placement:

Feed — ROAS 4.1x, CTR 1.8%, CPM ₹220. Verdict: carrying the campaign. Action: maintain, consider increasing reach via similar creative.

Reels — ROAS 1.9x, CTR 0.7%, CPM ₹165. Verdict: underperforming. Action: refresh creative with vertical-native format; current creative is 1:1 ratio repurposed.

Stories — ROAS 0.6x, CTR 0.4%, CPM ₹190. Verdict: dragging blended ROAS by 0.5x. Action: exclude this placement OR build Stories-native creative.

Click any card and Bach AI shows you the underlying creative-level breakdown.

When to Exclude a Placement vs Fix the Creative

Two different remediations for the same problem. Bach AI picks based on the data:

Exclude the placement when:

Placement-level ROAS is below 50% of campaign blended ROAS for 14+ days.
Creative is already format-optimized but performance still fails (the audience isn't there).
Spend on the underperforming placement is more than 15% of total — material drag.

Fix the creative when:

Format mismatch is obvious (1:1 in Reels, 9:16 in Feed).
Other creatives on the same placement perform fine (it's a creative problem, not a placement problem).
The audience profile suggests the placement should work.

Bach AI surfaces both options with predicted impact. You choose.

Why Indian D2C Accounts See Bigger Placement Gaps Than Global Accounts

Three reasons placement variance is larger in Indian accounts:

Reels dominance — Indian users spend 40-55% of Meta time on Reels, vs 25-35% globally. Reels CPM is lower but variance is higher.
Mobile-only audiences — Right Column and Marketplace are near-irrelevant. Spend efficiency is concentrated in Feed, Reels, Stories.
Creative localization — vertical Reels-native creative is rarer; many brands repurpose horizontal-shot Feed creative for Reels and pay the CTR penalty.

Net effect: the placement-level gap between best and worst placement is typically 3-5x in Indian accounts vs 1.5-2.5x in global accounts.

When Advantage+ Placements Is the Right Choice

Honest disclosure: turning off Advantage+ Placements isn't always the answer. Use Advantage+ when:

Your account is below ₹3L/month spend — manual placement breakouts get noisy at low volume.
You have placement-native creative for every placement.
Conversion volume is low enough that Meta's exploration helps.

Use manual placements when:

Your account is above ₹5L/month and one placement is clearly dragging.
Creative is format-specific (e.g., vertical Reels-only).
You want to allocate budget intentionally.

How Wittelsbach AI Manages Placement Decisions Across the Account

Bach AI runs placement-level analysis on every campaign continuously and surfaces exclusion or creative-refresh recommendations the moment a placement drags blended ROAS. No manual breakouts, no exporting to spreadsheets. Run a free Meta Ads audit at [app.wittelsbach.ai](https://app.wittelsbach.ai).

Frequently Asked Questions

Should I always exclude underperforming placements?

Not always. Some underperformance is creative-driven, not placement-driven. If your Reels ROAS is 1.5x but the only Reels creative running is a 1:1 repurpose, the fix is vertical creative, not exclusion. Bach AI distinguishes between these cases automatically.

Does excluding placements reduce reach significantly?

Excluding Stories typically drops reach 8-15%. Excluding Audience Network drops reach 5-12% with little ROAS impact (Audience Network is the lowest-quality placement for most D2C accounts). Excluding Reels would drop reach 30-40% and is almost never the right move in India.

How do I tell if a placement is dragging blended ROAS or just normally weak?

Compare to your account's blended ROAS, not a generic benchmark. If a placement is below 60% of your campaign's blended ROAS for 14+ days AND it's consuming over 10% of campaign spend, it's dragging meaningfully. Bach AI flags exactly this pattern.

Are Reels CPMs really lower in India?

Yes, typically 20-40% lower than Feed CPM for the same audience. But CTR is usually lower too, so cost-per-conversion can be similar or worse. The key metric is placement-level CPA, not placement-level CPM.

What if my placement breakdown shows mostly Marketplace or In-Stream spend?

Unusual for Indian D2C and usually a sign of auction starvation in the primary placements. Bach AI flags this pattern as 'budget escaping to low-quality placements' and recommends either creative refresh on the primary placements or explicit exclusion of the low-quality ones.