Why Your Facebook Ads Stopped Working (and How to Diagnose It)

When Facebook ads that were working suddenly stop, the cause is usually one of six things: broken tracking, a recent edit that reset learning, creative fatigue, a more expensive auction (often seasonal), a weaker offer, or a landing page problem. Diagnose in that order. First confirm the drop is real, then split it into cost to reach people, click-through and conversion rate, and fix the part that broke.

First: did performance really drop?

Before you touch a campaign, make sure you are looking at a real change and not a reporting artefact. A surprising number of "our ads died" moments are measurement problems.

1. Compare with your store. Did orders and revenue in Shopify or your store platform fall too? If store sales are steady but Ads Manager purchases fell, suspect tracking.
2. Check Events Manager. Is Purchase still arriving from both browser and server, at roughly the usual volume? A theme update, checkout change or app install can break it overnight.
3. Check the attribution setting. A change to the attribution window, or comparing columns with different settings, changes reported results without changing real ones.
4. Check the date range. Compare like with like: the last 7 days against the previous 7, not a weekend against a weekday.
5. Allow for delay. Recent days can look worse because some conversions are reported later. Judge the last two or three days cautiously.

If tracking is broken, fix that first. Nothing else in this guide can be judged until purchases are counted properly. Our Meta Pixel and Conversions API checklist covers the common breaks.

Where in the funnel did it break?

Every purchase from an ad is a chain: Meta buys impressions, people click, and some of them buy. Your cost per purchase is the product of three numbers:

CPA = CPM ÷ (1,000 × CTR × conversion rate)

So split the drop into its parts and compare each with the period when things worked:

Example: your CPA rose from $25 to about $39. CPM is flat, CTR fell from 1.4% to 0.9% and conversion rate held. That points to the ads, not the site. If CTR held but conversion rate fell from 3% to 1.8%, look at the site and the offer instead.

The ROAS calculator and CPA calculator help you see how much each change matters.

Triage flow: work through it in this order

1. Tracking: are purchases counted correctly? If no, fix tracking.
2. Recent changes: did you or a teammate edit anything in the week before the drop? If yes, consider that first.
3. Creative: is frequency rising and CTR falling on your main ads? If yes, refresh creative.
4. Auction: is CPM up across the board, including ads that didn't change? If yes, check season and competition.
5. Offer and site: is CTR steady but conversion rate down? If yes, look at the offer, price, stock and landing page.
6. Structure: are many ad sets competing, Learning limited or starved? If yes, consolidate.

Did a recent edit cause the drop?

Open the activity history for the campaign and look at the days just before performance changed. Meta's help page on significant edits lists changes that restart learning:

• Changes to targeting, creative, optimisation event or bid strategy
• Adding a new ad to an ad set
• Pausing an ad set for seven days or longer
• Large changes to the budget, bid control, cost per result goal, ROAS goal or spending limit

A drop that starts the day after you doubled a budget, swapped ads or changed the bid strategy is probably that edit. Ad sets in learning usually cost more. The fix is often patience, or reversing the change if it was a mistake. Avoid stacking more edits on top, which keeps resetting the clock. The learning phase guide explains what to expect.

Budget increases deserve special attention. Scaling too fast is one of the most common reasons a winning campaign "stops working": the extra money goes to less responsive people, and CPA rises. Our guide to scaling Meta ads without killing ROAS covers safer ways to grow.

Is it creative fatigue?

Ad fatigue is the most common cause of a gradual decline. The people most likely to respond have seen your ad, often several times, and the rest respond less.

Signs:

• Frequency rising week on week for the same ads
• CTR falling on ads that used to perform
• CPM steady or rising while results fall
• Comments repeating or turning negative
• Meta may show Creative limited or Creative fatigue in the Delivery column

The fix is new creative, but not random new creative. Keep the angle that worked and change the hook, the first frame, the format or the person on screen. Launch replacements in a batch, and pause tired ads only once a new one beats them. The full method is in ad fatigue on Meta ads, and the creative testing framework shows how to keep a pipeline of new winners.

Is the auction more expensive?

Sometimes nothing is wrong with your ads, and reaching people simply costs more. Signs: CPM up across most campaigns, including ads that didn't change, and CTR and conversion rate roughly steady.

Common reasons:

• Seasonal competition. Major shopping periods such as Black Friday, Cyber Monday, the end-of-year holidays and, in India, the festive season bring many more advertisers into the auction. CPMs usually rise in the run-up and often ease afterwards. See festive sale Meta ads in India.
• New competitors in your category bidding for the same people. The Meta Ad Library shows who is advertising now.
• A narrow audience that has been saturated, so remaining impressions cost more.

What to do: if your break-even still holds at the higher CPM, carry on and accept a lower margin for a few weeks. If it doesn't, cut budget to what remains profitable, lean on your strongest creative and offers, and hold back new tests until costs ease. Compare with the same period last year if you have the data.

Is it the offer, price or landing page?

If people still click at the usual rate but fewer buy, the ads are doing their job and something after the click has changed.

Check:

• Price and offer. Did a discount end, a price rise, or free shipping move to a higher threshold? Ads promising the old offer now meet a worse one.
• Competitor offers. A rival running a big sale can pull buyers away even if your page hasn't changed.
• Stock. Popular sizes or variants out of stock quietly kill conversion.
• Site speed and errors. Test the page on a phone over a mobile connection. Check checkout works, including payment methods.
• Message match. Does the page show what the ad promised? A new homepage design can break the link between ad and page.
• Reviews and trust. A run of negative reviews or comments under the ad can lower conversion.

Conversion rate problems are usually fixed on the site, not in Ads Manager.

Is your campaign structure holding results back?

Structure problems usually build slowly rather than appearing overnight, but they show up when you scale.

• Too many ad sets each getting too few purchases to learn, often showing Learning limited.
• Overlapping campaigns, such as an Advantage+ sales campaign and a manual broad campaign using the same ads, competing for the same people.
• Retargeting soaking up spend under one campaign budget, flattering short-term results while new-customer growth slows.

Consolidate into fewer ad sets with distinct jobs. The CBO vs ABO guide shows how to size a structure to your budget.

Is the drop real for the business?

Finally, step back from Ads Manager. Look at total store revenue against total ad spend, your MER, over four to eight weeks.

• Ads Manager ROAS down, MER steady: attribution may be shifting credit, for example to other channels or to organic. Your ads may be working as well as before.
• Ads Manager ROAS steady, MER down: ads may be taking credit for sales that would have happened anyway, often from existing customers.
• Both down: a real decline. Work through the triage flow above.

How do you stop it happening again?

Most sudden drops are easier to catch early than to fix late. A short weekly routine helps:

• Review the same numbers every week: spend, purchases, cost per purchase, CPM, CTR, conversion rate and frequency for your main campaigns, against the previous week. Our weekly report guide has a template.
• Keep a change log. Note every budget change, new ad batch and setting change with the date. When results move, you'll know what changed.
• Keep new creative in the pipeline, so a tired ad can be replaced in days, not weeks.
• Run a monthly tracking check: a test order, a look at Events Manager and a comparison of store orders with reported purchases.

What not to do when ads stop working

1. Don't rebuild everything at once. New campaigns, new audiences and new ads together mean you never learn what fixed it.
2. Don't make daily edits. Each significant edit restarts learning.
3. Don't panic on two bad days. Daily results swing. Look at seven days against the previous seven.
4. Don't turn off a whole account of winners because one metric moved. Pause losers, protect winners.
5. Don't assume the algorithm is broken. It is almost always one of the causes above.

ROAS drop diagnostic checklist

• Store sales compared with Ads Manager purchases
• Purchase event checked in Events Manager, browser and server
• Attribution settings and date ranges consistent
• CPM, CTR, conversion rate and frequency compared with the good period
• Activity history checked for edits before the drop
• Creative age and frequency reviewed, refresh batch planned
• Season and competitor activity checked
• Offer, price, stock and landing page tested on a phone
• Structure checked for overlap and Learning limited ad sets
• MER reviewed over four to eight weeks

Keep a steady hand

Fly ROAS helps with the parts of this that are easiest to neglect: it studies long-running competitor ads, builds fresh on-brand creatives from your real product photos so tired ads can be replaced quickly, and its Autopilot judges every test after 48 hours inside the caps you set, explaining each decision in plain English.

Diagnose before you fix

A ROAS drop feels urgent, which is exactly why it is worth slowing down. Confirm the drop is real, split it into reach cost, clicks and conversion, check your own recent changes, then look at creative, the auction, the offer and the page, in that order. One clear cause, one clear fix, measured over a week, beats a dozen rushed changes every time. If your ads have stopped spending altogether, start with Facebook ads not spending instead.