Facebook Ads Not Converting? A Step-by-Step Diagnosis Checklist

Your Facebook ads are spending money, people are clicking, but sales aren't coming in. Or worse: the campaigns worked fine last month and suddenly stopped converting. You're not alone. Reddit's ad communities are full of posts like this one: a skincare brand spent $2,300 over six weeks and got 4 sales on a $135 product, despite a decent click-through rate. Another store owner found that AddToCart and Purchase events were simply not arriving in Meta. A fashion brand was paying a $556 CPM because the targeting was cut down to a few postcodes and age brackets.

Each looks the same from the outside, but each needs a different fix. This guide gives you a diagnosis order: start with what is cheap to check and breaks most often, then move to the bigger strategic questions.

Step 0: Define which problem you actually have

Before you change anything, answer one question: did these ads never convert, or did they stop converting?

• Never converted: Usually a problem with tracking, the optimization goal, the offer, or the landing page.
• Suddenly stopped: Usually something broke (tracking, website, checkout, payment), a Meta-side delivery issue, or creative fatigue.

Write down when things changed and what happened around it: theme updates, new apps, price changes, edits to winning ads.

Step 1: Check that tracking works (most common culprit)

If Meta can't see your conversions, it can't optimize for them, and you can't judge your ads. Check these points in order:

• Compare numbers: Put your shop or CRM numbers next to what Ads Manager reports for the same period. A big gap in either direction means tracking is off. Some advertisers see fewer conversions than they really have, others see more (for example, when a lead event fires on page load instead of on form submit).
• Test events: Open Events Manager, use the "Test events" tool, and walk through your own funnel: product page, add to cart, checkout, purchase. Every step should show up once.
• Duplicate or outdated pixel code: On Shopify, a classic problem is old pixel code pasted into the theme years ago, running alongside the official Facebook and Instagram sales channel app. Remove the manual code and let the integration handle events.
• Conversions API: Browser tracking alone misses events because of ad blockers and privacy settings. Set up the Conversions API (most shop systems offer it through their official integration) and check event match quality in Events Manager.
• Right event, right place: Make sure the campaign optimizes for the event that actually matters (Purchase or Lead), and that this event has fired recently.

Don't touch anything else until tracking is clean. Every other step depends on it.

Step 2: Rule out outages and account problems

If performance dropped overnight across all campaigns, check whether it is just you. Meta delivery issues do happen, and when they do, ad communities on Reddit usually light up within hours. Also check Account Quality in Business Settings for restrictions, rejected ads, or payment problems. If Meta confirms a delivery bug, document the affected days.

Step 3: Check what the campaign optimizes for

Meta delivers exactly what you ask for. With a Traffic or Engagement objective, it finds people who click or like, not people who buy. That alone explains many "lots of clicks, no sales" accounts.

• Use the Sales objective with the Purchase event (or Leads with a lead event) once tracking works.
• Look at placements. If most clicks come from Audience Network, many of them are likely accidental taps. Check the placement breakdown in Ads Manager.
• If you have very few purchases, consider optimizing temporarily for an earlier event like Add to Cart or Initiate Checkout, and switch back to Purchase once you have more volume.

Step 4: Find where people drop off

"Not converting" is too vague to fix. Break the funnel into stages and find the leak. In Ads Manager, add the columns for link clicks, landing page views, adds to cart, checkouts initiated and purchases. Then look at the drop between each stage:

• Clicks to landing page views: A big gap points to slow loading, broken links, or accidental clicks.
• Landing page views to add to cart: People arrive but do not want the product at that price. This is an offer, product page, or message mismatch problem.
• Add to cart to checkout: Surprise shipping costs, forced account creation, or missing trust signals.
• Checkout to purchase: Payment options, errors in the checkout, or doubts at the last step. Place a test order yourself on your phone.

Once you know the stage, you know which part to fix. A better ad will not fix a broken checkout.

Step 5: Look at CPM and targeting

CPM (cost per 1,000 impressions) tells you how expensive it is to reach your audience. A high CPM is not automatically a problem, but an extreme one usually signals over-restricted targeting. When you narrow your audience to a few postcodes, an age range and several interests, Meta has very few people left to show your ads to and has to pay more to reach them.

• Test a broad audience (country, age range, gender only if truly needed) and let your creative do the targeting.
• Compare CPM by ad. Ads with a strong hook often get cheaper delivery because people engage with them.
• Remember that CPMs usually rise in competitive periods like Q4. Judge results by cost per purchase, not by CPM alone.

Step 6: Check creative fatigue and variety

If an ad worked and then faded, look at frequency (how often the same person saw the ad) and click-through rate over time. Rising frequency and falling CTR mean your audience has seen it enough.

The fix is not ten versions of the same video with different music. Meta's delivery system treats very similar ads almost like one ad. What works is different concepts: a customer testimonial, a problem and solution demo, an answer to the most common objection, a founder story. Plan a few new concepts every month and keep the winners running untouched, because edits to a running ad can reset its learning.

Step 7: Be honest about the offer and landing page

Take the skincare example: a $135 product sold to women 45+ through a cold ad is a big ask. Good CTR shows interest, poor sales show missing trust or a weak offer. Ask yourself:

• Does the landing page continue the exact promise of the ad, or does it land on a generic homepage?
• Are there real reviews, before and after results (if your industry allows it), guarantees and clear return policies?
• Is there an easier first step, such as a starter size, bundle, or a lead magnet for higher-priced products?
• Does the page load fast on mobile, where most Meta traffic comes from?

Step 8: Match budget and structure

Meta's algorithm needs data. As a guideline, Meta says an ad set needs around 50 optimization events within seven days to exit the learning phase. If your budget is small and spread across many campaigns and ad sets, none of them will get there, and results stay unstable.

• Consolidate into fewer campaigns with more budget each.
• Avoid big budget jumps. Increase in smaller steps so the campaign does not restart learning.
• Give changes a few days before judging them, unless something is clearly broken.

If you are unsure how much budget your goal requires, our guide on calculating your ad budget walks you through it, the ad cost calculator for business owners helps you check your numbers, and ad costs and contribution margin shows how much a sale can cost you before it stops being profitable.

What's in it for you

You're paying for ads every day, and every day without sales feels like money down the drain. You don't want to guess, and you don't want to rebuild everything for nothing.

With a fixed diagnosis order, you find the real leak instead of changing ten things at once.

What you get:

• Less wasted spend. You fix the cheap, common causes (tracking, objective, checkout) before paying for new creatives or a new agency.
• Clarity, fast. Each step tells you where to look next, so you stop second-guessing every dip.
• Calm decisions under pressure. When results drop, you know whether to wait, fix or stop.
• Data you can trust. Once tracking is clean, every future test tells you something real.

The checklist won't turn a weak offer into a strong one. It will show you quickly whether the ads, the page or the offer is holding you back.

FAQ: specific situations

Should I wait for the learning phase or stop?

Separate funnel problems from learning. If clicks barely turn into installs, add to carts or signups, learning will not fix that part. If the funnel looks healthy and you simply have few conversions yet, give it more data. Set a kill rule before you start, for example pausing an ad once it has spent one to two times your target cost per purchase without a single sale.

Will a new pixel, a new ad account or a pixel reset fix it?

Rarely for long. If every new product does well on day one and then collapses, or a pixel reset helps only briefly, the cause is usually creative fatigue, a small audience or the offer. A fresh pixel throws away the data Meta learned from. Only consider a new account if Meta has actually restricted the current one.

People open the lead form, but nobody submits

If completions dropped to zero suddenly, something probably broke. Submit the form yourself on a phone, check required fields and confirm the lead event still fires. If completions were always low, shorten the form or offer an easier first step.

I get messages or leads, but they are spam or unqualified

Messages and engagement goals attract people who like to chat, not buyers. Add one or two qualifying questions, mention the price range in the ad, and send real outcomes (booked appointments, purchases) back to Meta so it learns what a good lead looks like.

Meta reports more leads than I actually have

Then the event fires at the wrong moment, for example on page load or quiz start. Move it to the real completion step and compare with your CRM or database every week.

My product is high ticket. Does the structure matter?

Yes. Ten creatives each in their own ad set split a small budget so thin that none gets enough data. Group creatives in fewer ad sets, judge on add to carts and checkouts first, and give buyers more proof and time to decide.

Good CTR, but the cost per sale is far too high

The ad attracts attention, but the page or offer does not convince. Check the funnel stages in Step 4, compare the price with competitors and look at what visitors see in the first seconds on mobile.

Meta confirmed an outage. Can I get money back?

When Meta acknowledges a delivery issue, contact support with the affected campaigns and dates and ask about ad credits. Document everything, and do not rebuild your account during the outage.

Conclusion: diagnose first, then fix

When Facebook ads are not converting, resist the urge to change everything at once. Go through the steps in order. Most accounts have one main leak, and once you find it, the fix is usually clear.

If you'd rather have experts handle this: at Plaith, 6 specialists, combining people and AI, take care of ad videos, ad messaging and campaign management for your Meta ads, from €26 a day plus a one-time setup. If you want an honest second opinion on why your ads are not converting, book a short demo call here.