Why are my Facebook ads not spending when my inventory is already paid for?

An ad set that refuses to deliver is a cash problem before it is a marketing problem. The units were bought, the supplier was paid, and the shelf space is occupied. Every flat day the money sitting in those units earns nothing and the window you bought them for gets shorter.

The technical sequence below comes from one source video, AdAmigo AI's walkthrough for ad sets that will not deliver. Everything about unit margin, catalog structure, season windows and unsold stock is this site's application to shops that own what they sell. The video does not cover any of it.

Why is my Facebook ad set not spending the full daily budget?

Four checks, in the order the source video runs them: account standing and payment method, ads inside the set that are off or rejected, audience size at the ad set level, and a manual bid set too low to win auctions. Calling that order cheapest first is this site's framing, not the video's.

The first two checks cost nothing to run and change nothing in the account. The last two change how the set enters the auction, so neither can be judged the moment you save it.

AdAmigo AI frames the trigger plainly. When a set with a daily budget of 240 ends up spending close to nothing, in the channel's words, "Then we know something is not right and we need to take action."

What the video does not weigh, and what changes the whole calculation for a shop holding goods, is that the clock is not neutral. Work out these two numbers before you start debugging, using your own figures and nobody else's:

• Contribution margin per unit. Selling price minus landed cost, minus payment fees, minus outbound shipping and packaging. This is the number your acquisition cost has to fit under, not the sticker price.
• Days of cover left in the window. How many selling days remain before the line goes to markdown, gets replaced by next season, or ties up space you need for incoming goods.

A dead ad set early in a season window is an annoyance. The same dead ad set with only a short stretch of full price selling left is a markdown decision that has already been made for you.

What should I check first when an ad set stops delivering?

Confirm the account is in good standing and the payment method cleared, then open the ad level and confirm every creative is active and not rejected. If other sets in the same account spent money over the same period, the account is healthy and the fault lives inside the one set that stalled.

AdAmigo AI describes the fastest version of that test: "The first hint for that is if you have other audiences that are spending money, right?"

That test is unusually cheap for anyone running a wide product range, because the account already contains a natural control group. Sets pointed at other categories are running under the same billing, the same pixel and the same page. If those are spending and one is not, you have isolated the problem without changing anything. The video does not make that point about multi category accounts. It is this site's reading of the same test.

Then go one level down. The channel notes that ad level problems roll upward: "So, if all of your ads are disabled or blocked for whatever reason, then your adsets won't deliver as well."

Three things worth doing at the ad level before you go further:

1. Turn on the delivery column. The video walks the path through customize columns, 3:09. Without that column you are guessing which creative is blocked.
2. Check each ad individually, not the set total. A creative library built per product line grows one asset at a time, and each of those assets can be rejected on its own.
3. Check the landing destination. If the product page behind the ad is out of stock or unpublished, you are debugging delivery on an ad pointed at nothing. That check is not in the video. It belongs to anyone whose page can sell out.

How large does my audience need to be for ads to deliver?

Audience size sits at the ad set level, and narrow layered interests or a thin retargeting pool are the usual causes. AdAmigo AI puts the caution line at 200,000 people, and flags website custom audiences built on a pixel with fewer than a thousand visitors. Widen the targeting or split layered interests into separate sets.

The channel is direct about the threshold: "What's too narrow? I would say that anything below 200,000 is playing with fire." And on the retargeting side, "Also when you target a custom audience a retargeting audience of your website visitors let's say and on that pixel you have less than a thousand visitors that this could also be causing that issue."

Here is where a wide product range works against you, and the video never raises it. This section is this site's application.

The instinct when you carry many lines is to build one ad set per line, or worse, one per item. It feels like control. Each of those sets then gets its own tight interest stack so the targeting "matches" the product. The result is a structure that can leave every set sitting under the threshold the channel warns about.

The same thing happens on retargeting. A viewers audience for one product page is a fraction of your site traffic. Build one per page and you end up with a row of pools that are individually too thin to run against, even though the combined traffic would be fine.

Practical way out, following the channel's own fix of unlayering interests into separate ad sets:

The trade you are making is real. Merging sets means you lose per item read on the numbers. A set that delivers and reports at category level is still worth more than a shelf of sets that report nothing because none of them spent.

There is a second possibility the video does not mention, and it only comes up when many sets point at overlapping buyers. Near identical sets inside one account may end up competing for the same impressions. If you have already fragmented your structure, some of the delivery problem may be self inflicted. Treat that as a hypothesis to test by consolidating, not as a claim about how the auction works.

Can a low manual bid stop my ads from spending?

Yes. A cost per result value set too low is one of the causes the source video lists, and it names removing the cap as the first move. Shops that price acquisition straight off a thin unit margin are the ones most likely to type in a number the auction never clears.

The channel's account of what a low value does: "you don't get any any reach because you're never winning an auction, right?" The example it uses is a $5 cost per purchase target on a product priced far above that, and the conclusion it draws: "So, um it always has to be in relation." The fix offered is blunt: "So, my recommendation would be in the beginning to just get rid of it, to not use it, um and see if that fixes the issue."

Now the part the video does not address. Anyone selling goods at retail markup arrives at that value from the wrong direction. The reasoning goes like this: my contribution margin per unit is X, so I cannot pay more than X to acquire a sale, so I will type X into the cost per result field and let the platform sort it out.

That is arithmetic, and on the channel's own account it is also how a set ends up never winning. Two corrections worth making before you type a number:

• The value is what you tell the platform you are willing to pay, not what you are promised. In the channel's framing it always has to be in relation to what you sell. Set it at your margin ceiling and, by that same account, silence is a possible outcome rather than a cheaper result.
• Your break even acquisition cost is not your first purchase margin alone. If you have repeat buyers, replenishment lines, or bundles that move together, the number you can defend is higher than a single unit calculation suggests. Work it out from your own repeat rate. Do not borrow a figure from anywhere.

If your unit economics genuinely cannot support the clearing price of the auction, that is a pricing or assortment answer, not a bid setting answer. Turning the value off will tell you what the auction actually costs, and then you can decide whether that line can carry it.

How do I force a stalled ad set to start spending?

Three escalating moves, in the source video's order: duplicate the set so it goes back into the auction, then set a deliberately high manual bid, then raise the daily budget and switch delivery type to accelerated where that option exists. The channel moves on after half a day to a day or two without spending, then after another day.

Step one, duplicate. The channel's reasoning is that a fresh entry is sometimes enough on its own: "Just re-entering the marketplace for whatever reason gets the spend going."

Before you duplicate, check that the item behind the creative is still available in the sizes or variants the ad shows. Paying for a fresh entry on an ad that points at a sold through variant is a way to buy traffic you cannot fulfil. The video does not raise this. It applies to anyone whose stock is finite.

Step two, high manual bid. This is the counterintuitive one, and the channel says so itself: "I would bid very aggressively. I get sometimes attacked for this because it's not a really logical thing to do, right?" The worked example opens in that same block with a $500 product and finishes in the next one, with a target cost per purchase of 250 and a bid of 1,500: "So, three times your AOV, two times your AOV, something very aggressive because I want to reenter the auction."

The channel is explicit that the bid is not the cost it expects to pay: "And it won't really translate into 1,500 cost per purchase." On why it is there at all, "And we're just using this as an initial kick in the butt to get some spend going." You take it back out once the set is delivering.

This site's addition, which the video does not make: run that multiple against your own average order value and your own contribution margin, and decide in advance the total amount you are willing to lose during the kick start period. Write the number down before you change the bid, not after. If a spike does land at a cost close to your bid, on a line with retail level margin, the loss is absorbed per unit and it compounds fast.

Step three, budget and accelerated delivery. The channel goes to "which is raising the daily budget and if possible, accelerated spending." The path shown is inside the ad set, under show more settings, then delivery type. It is not available in every account, and the channel says to check for yourself. Its own description of the effect: "And so then it will basically be yet another kick in the butt to really accelerate spending and get this going."

The warning attached to it is the important line: "But make sure that you monitor every hour to ensure that you know you're not um you're not burning through capital too fast."

Hourly monitoring means something extra when the goods are finite. Accelerated delivery paired with a raised budget can outrun availability on a fast moving variant, and an oversell on a line you cannot restock inside the season is a refund queue plus a support load plus the ad spend you already paid. Set your low stock threshold before you switch it on. That instruction is this site's, not the video's.

When should I stop fixing an ad set and move the budget somewhere else?

Kill it when the cost of the fix passes the value of what the set could still sell. Two numbers decide it: contribution margin per unit on the line being advertised, and selling days left in the window. AdAmigo AI's exit rule is simpler, which is to lean on the audiences that already deliver.

The channel's version: "if you have other audiences that are delivering and performing well, you could obviously also just throw in the towel at some point"

That is sound advice for accounts where all budget is fungible. It is incomplete when a specific set exists to move a specific batch of goods you already own. The framing below is this site's, and the video does not sell it.

Ask three questions in order:

1. Is this set the only route to this stock? If the same units sell through another delivery channel, another audience, or a bundle that is already moving, shut the set and reallocate. The stock problem is solved without it.
2. How many selling days are left? The escalation sequence above is not instant. The channel's own pacing is half a day to a day or two before the second step, and another day before the third. Count that against the selling days left on the line. If the debug cannot finish inside the window, go and plan the markdown instead.
3. What does the fix cost against the margin at stake? Total contribution available on the remaining units is your ceiling. If the aggressive bid phase plus the accelerated delivery phase could plausibly consume a meaningful share of that ceiling, the fix is not worth running. Compute both sides with your own figures.

There is a fourth answer that is not really a marketing answer. Sometimes the honest read is that the goods were bought wrong, and no delivery fix changes that. Clearing them at a controlled discount and freeing the working capital beats spending another week trying to force delivery on a set that is being ignored for a reason.

Which tools help diagnose and run ads for an online shop?

Different jobs sit in different places. Auction diagnostics live inside the ad platforms themselves. Creative production, list marketing and conversation handling sit in separate tools and do not touch delivery settings. The table below describes what each one does and what it assumes you already know.

SaleADS.ai is the product of the company that publishes this site. It appears in the table for comparison, not as a recommendation.

Where does this information come from?

Every ad platform step described above comes from one video by AdAmigo AI, linked at the block where each step is spoken. Everything about unit margin, catalog structure, stock cover and markdown timing is this site's application to shops that own what they sell, and it is listed separately below so that the two are never confused.

Video: "Facebook Ads Not Spending? Step-By-Step Fix for Ad Sets That Won’t Deliver (Meta & Instagram)" Channel: AdAmigo AI Link: https://www.youtube.com/watch?v=ubl-MFEWxmQ

What comes from the video

• The framing that a large gap between daily budget and amount spent means action is needed, 0:00
• Checking account standing, payment method and warning messages first, and using other spending audiences as the signal that the account is fine, 0:47
• The point that disabled or blocked ads stop the set from delivering, 1:34
• The 200,000 audience caution line, the advice to stop layering interests and split them across separate ad sets, and the note on custom audiences built on a pixel with fewer than a thousand visitors, 2:22
• Waiting to accumulate more visitor data before relaunching a thin retargeting audience, and enabling the delivery column through customize columns, 3:09
• The low manual bid problem, the $5 example, the statement that you get no reach because you never win an auction, and the recommendation to remove the value and see whether that fixes delivery, 3:57
• Duplicating the campaign or ad set as the first force spend move, the wait of more than half a day to a day or two before escalating, and the opening of the $500 product example, 4:46
• The 250 target cost per purchase and the 1,500 bid that complete that example, the two to three times average order value framing, and the statement that the bid will not translate into the actual cost per purchase, 5:36
• Raising the daily budget and using accelerated delivery type where available, found under show more settings, 6:24
• Monitoring hourly to avoid burning capital too fast, and the option to abandon the set and rely on audiences that already perform, 7:11

What is this site's application and is not in the video

• Treating a stalled set as a cash problem tied to goods already paid for, and the two numbers to compute first, contribution margin per unit and selling days left in the window
• Reading the video's check order as cheapest first
• Using other categories in the same account as a built in control group when testing account health
• Checking that the product page and variants behind the ad are still available before debugging delivery
• The catalog fragmentation trap, where one set per item plus layered interests pushes every audience under the threshold the video warns about
• The suggestion that near identical sets pointed at overlapping buyers may compete with each other, offered as a hypothesis to test by consolidating
• The reason owners who price acquisition straight off a thin unit margin end up with a value that never wins
• Writing down a maximum absorbable loss before switching to an aggressive bid
• Setting a low stock threshold before enabling accelerated delivery, because oversell risk applies when supply is finite
• The three question exit test, and the point that sometimes the correct answer is a controlled markdown rather than another week of delivery debugging