Meta Ad Placements: Where Your Ads Run and When to Trim

Every Meta campaign has a placements setting, and many small businesses never open it. Left alone, Meta decides where each rupee goes: a Facebook feed, an Instagram Reel, WhatsApp Status, a search result, or an app that has nothing to do with Meta. Owners who do look often spot one expensive-looking line and switch it off.

Neither habit is a plan. What usually holds up for a small service business is the default, fed with creative that fits each screen, and checked against real enquiries before anything gets cut.

Where a Meta ad can show up

Meta's current list spans Facebook, Instagram, Messenger, WhatsApp, Threads and Audience Network, though not every placement is open to every objective or format.

• Feeds: Facebook, Instagram and Threads feeds, Facebook Marketplace, Instagram's Explore grid, Facebook notifications and the Facebook right column (computers only).
• Stories, status and reels: Stories on Facebook, Instagram and Messenger, Reels on Facebook and Instagram, and WhatsApp Status.
• In-stream: ads inside Facebook videos and while people watch Facebook Reels.
• Search: next to Facebook and Marketplace search results, and in Instagram search.
• Apps: Audience Network, which shows your ad inside other publishers' apps.

Meta says adding placements does not raise the cost of your ad.

What Advantage+ placements actually does

Advantage+ placements, which older guides call automatic placements, lets Meta spread one budget across every placement your ad qualifies for. It is the default, and Meta recommends it because the delivery system can then try to make the most of your budget.

Under the highest volume bid strategy, Meta says the system aims for the most results at the lowest average cost overall, rather than the lowest cost in each placement. Its help page gives an example where Instagram costs more per result than Facebook, yet switching Instagram off leaves fewer results at a higher average cost. A placement that looks expensive on its own row can still be earning its place.

The gap, for a service business, is that Meta optimises for the event you chose, usually a form submission or a started conversation. Unless you send call outcomes back (the lead quality piece covers how), it can't tell who answered your team's call, so a placement full of cheap, silent form fills keeps getting budget. That is the main reason to override the default, and only your own lead sheet can prove it.

Manual placements, and what changed for lead campaigns

On Awareness, Traffic and Engagement campaigns you can still choose by hand: in the ad set's Placements section, edit Advantage+ placements, pick Manual placements and untick what you don't want.

For Leads, Sales and App promotion, Meta is gradually removing placement choice inside the ad set. Your account may still show Placement controls under Show more settings, or it may be gone. Two controls matter either way:

• Account-level placement controls. In Advertising settings, under Placement controls, you can exclude Audience Network, Facebook Marketplace and the Facebook right column. These apply to every campaign in the ad account, Advantage+ placements included. Meta is still rolling this control out.
• Limited spend on excluded placements. Exclude a placement inside an ad set and eligible accounts are automatically opted in to Meta spending around 5% of the budget on each excluded placement when it expects better results. For zero spend, turn it off under Manage excluded placements.

Read the placement breakdown before you change anything

In Ads Manager, open the Ad sets tab and click Breakdown, the drop-down between Columns and Reports. Under By delivery, pick Placement, or Platform for a Facebook-versus-Instagram split. Meta labels these figures as estimates.

Judge cost per result, not click-through rate, because some placements collect plenty of taps and few enquiries. Ignore rows with a handful of results: three leads at ₹900 against forty at ₹420 is too thin to judge. A sensible floor is a few dozen results per placement before calling it worse.

Then match rows to real outcomes. If enquiries come through your website, add Meta's dynamic URL parameter placement={{placement}} in the ad's URL parameters field before launch (Meta notes that changing them later may restart learning). Your analytics, or a form that captures the page address, can then record each visitor's placement. Meta fills it only for placements on its supported list, so some visits won't be labelled. If your Pixel or Conversions API reports a deeper website event, such as a booked consultation, add it as a column first.

Instant forms are harder, because your calling sheet doesn't know which placement each lead came from. There, run a before-and-after test: record your answer rate for a few weeks, then exclude the suspect placement and compare.

Audience Network, the placement people argue about

Audience Network puts your ad inside apps run by other publishers, who earn money by showing ads from Meta advertisers.

Some advertisers exclude it for lead generation, and the worry is reasonable: a tap inside an app can be less deliberate. Meta's own guidance for app publishers discusses accidental clicks, such as clicks that register very soon after an ad appears. That doesn't prove it's worse for your account, but it makes it the first row to check, on answered leads rather than form counts. If it clearly loses, the account-level control is the cleanest cut.

Give each placement creative that fits

Plenty of placement trouble is really creative trouble. Missing shapes get filled by padding, cropping or expanding your media, and an ad won't deliver where its media isn't supported, so a weak Reels row may never have had a proper vertical ad. Meta's recommended shapes:

• Feeds: square 1:1 or vertical 4:5. Meta suggests 4:5 for single images in Facebook Feed and 1:1 in Instagram Feed.
• Stories, Status and Reels: full-screen vertical 9:16, which Meta also recommends for Audience Network.
• In-stream video: horizontal 16:9.

At minimum, make a 9:16 and a 4:5 or square version of every ad, then set the crop or text for each placement with asset customisation in Ads Manager (media customisation in the newer workflow). Turn on the Safe Zone guardrail while you check, and keep text and logos away from every edge of 9:16 creative, since the profile icon or call-to-action button can cover them.

Cropping an asset for one placement doesn't stop it running elsewhere; opt it out too. And Meta treats adding a new ad or changing the creative as a significant edit that restarts learning, so build the vertical version before launch. The creative piece covered what the ad itself should say.

When to trim, and how

1. Launch on Advantage+ placements with a vertical and a feed-shaped version of each ad.
2. Leave placements alone for two to three weeks, longer if results are thin.
3. Set each placement's spend against real results: an answered lead, a call that reached someone, a booking.
4. Exclude a placement only when its cost per real result is clearly worse over a decent sample. Slightly worse, or worse on six leads, doesn't count.
5. Make one exclusion at a time and re-check the whole ad set two to three weeks later. If total cost per real result got worse, put the placement back.

Take a hypothetical coaching centre spending ₹30,000 over three weeks, with a website form that records the placement parameter. Audience Network takes ₹6,000 and brings 60 enquiries; everything else takes ₹24,000 and brings 140. Per enquiry, Audience Network looks cheaper at ₹100 against about ₹171. But the counsellors' sheet shows 5 of its enquiries answered against 70 of the rest: ₹1,200 per answered lead against about ₹343. That gap, on that many leads, is worth acting on; a difference of two or three answered leads would not be.

FAQ

Do I need an Instagram account to run ads on Instagram placements?

No. Meta lets your Facebook Page act as the voice of the ad on Instagram. A real profile still helps when people tap your name.

Should I switch off Audience Network for lead forms on day one?

Only with a hard reason, such as a brand or policy constraint. Otherwise let it run, compare its cost per answered lead after a few weeks, and if it is clearly worse, exclude it in account controls.

Why does my report still show spend on a placement I excluded?

Check whether Allow limited spend to excluded placements is switched on for that ad set. Account-level controls can also take up to 48 hours to reach running campaigns.