Of its Audience Network filtering, Meta says "it is unlikely that all invalid activity can be identified and removed" (Audience Network methodology), so bots are a real branch. But in most store-owner threads we read, the replies pointed to tracking, offer or page problems, not fraud. This is the diagnostic step inside our guide to Meta Ads click fraud.
How we checked this. No ad-account test was run. Sources are Meta's help and developer pages read on 17 November 2026, and store-owner threads from 2022 to 2026 on a store platform's community forum, quoted as reported.
If the conversions you want are leads, not purchases, and your Facebook ads get clicks but no conversions, the problem is usually lead quality: read why Advantage+ leads do not turn into customers, and validate Instant Form leads in your CRM.
The six checks, in order, with the Meta fact behind each:
- Step 0, tracking and objective. Ads Manager and your store count purchases differently by design, and a Traffic campaign is set to drive visits, not sales.
- Step 1, page loads. Meta counts a landing page view only when the page loads after the click, seen through its in-app browser or your Meta Pixel. Far fewer views than outbound clicks means sales are lost before the page.
- Step 2, placement and device. Meta's offsite-conversion breakdowns include Placement, Platform and Impression device, so read purchases per row, not clicks.
- Step 3, hour and frequency. Neither of Meta's breakdown lists for offsite conversions includes Time of day, so set store orders by hour against clicks by hour, over at least a full week.
- Step 4, the funnel drop. Read ViewContent, AddToCart, InitiateCheckout and Purchase in order. Meta says leaving the learning phase "usually occurs after about 50 results in the week after the ad set's last significant edit".
- Step 5, invalid traffic, last. Only when a shortfall holds in one placement, device or hour band. Meta does not charge for clicks its manual review finds invalid, and it usually bills for impressions, not clicks (How Meta Charges for Ads).
Step 0: Are there no sales, or no sales Meta can see?
Compare store orders with Ads Manager purchases for the same dates first, because Meta's counts can differ from your store's by design. Clicks with no purchases in Meta are not always clicks with no orders.
- Store has orders, Ads Manager shows fewer. Meta's checklist names the "Attribution window", "Time zone" and "Conversion date", and says: "Check whether the raw pixel fires in Events Manager match your third-party reporting tool." (Meta's reporting differences checklist).
- Attribution setting. Meta's standard click-through setting counts events "within 1-day or 7-day after a link click on your ad", and view-through "within 1-day after an impression of your ad" (About attribution settings).
- Purchase events arriving. For Conversions API events, Meta says you "should be able to verify them within 20 minutes after they were sent" in Events Manager (Verifying Your Setup).
- Ads Manager shows more purchases than the store. The same checklist suggests the Meta Ads Data Advisor "to check for duplicate pixel fires, which can sometimes be caused by redirects." Purchases with no orders at all are the reverse symptom: see fake pixel events.
- Uncategorized rows. "When using breakdowns for conversions, all conversions that we aren't able to attribute to, will show up as 'Uncategorized.'" (breakdowns in Meta Ads Reporting).
- Objective and performance goal. Meta describes Traffic as "Increase traffic to an online destination of your choice" and Sales as "Find people likely to purchase your goods or services" (How to choose an ad objective). A Traffic campaign working as set can bring clicks without sales. In a Sales campaign, check that the ad set's performance goal and conversion event name a purchase, not link clicks or landing page views, because delivery seeks the result you choose. When Meta predicts zero purchases it suggests a more common event (Meta's zero-conversions guide), which moves optimization before the purchase.
If tracking is clean and the campaign optimizes for a sales event, go to Step 1.
Step 1: Do the clicks load your page?
If landing page views run far below outbound clicks, the sales are lost before the page. One of Meta's reasons landing page views run lower: "The web page isn't fully loading or the app failed to open after a link click." (Meta's link clicks and landing page views page).
If so, run the link clicks vs landing page views gap test. If both counts look normal, you have Facebook ads traffic but no sales from visits that load: go to Step 2.
Step 2: Which placement and device get clicks but no sales?
Read purchases and cost per result per placement, platform and device, not clicks.
- Know the purchase limit. Under By delivery, the offsite-conversion list in Breakdowns in Ads Manager is "Age, Gender, Age and gender, Country, Impression device, Platform, Platform & device, and Placement." Country is in that list: a country you do not sell to that still takes clicks is worth a look.
- Read Audience Network as one row. Meta's Audience Network methodology lists its accredited placements as native, banner, interstitial, Connected TV in-stream video and rewarded video units. With Advantage+ placements, Meta can deliver there unless you exclude it, and a costlier row is not proof of waste. If that row holds the clicks and none of the purchases, see how to exclude Audience Network on Meta ads and our page on Audience Network fraud.
- Check the in-app browser. In the Facebook and Instagram apps, "web links in the ad open in the in-app browser by default" (About the in-app browser). Our reading: switching browsers, or returning later, can break the session your store sees, so sales look missing with no fraud. The in-app browser is not Audience Network.
- Use link clicks, not Clicks (all). Clicks (all) "includes link clicks as well as clicks on other parts of your ad" (Clicks (all) vs Link clicks). CTR (link click-through rate) is "link clicks divided by impressions" (CTR (link click-through rate)). It measures the ad, not the buyer, so a high CTR with low conversion on Facebook ads is no contradiction.
If one row holds the clicks and none of the purchases, note it for Step 5. If the shortfall is spread across rows, go to Step 3.
Step 3: Do the clicks bunch by hour, or is frequency rising?
Set your store's orders by hour against Meta's clicks by hour, and read rising frequency with falling results as fatigue, not fraud.
- The hourly path. Choose By delivery, then a Time of day breakdown (View hourly breakdowns).
- Do not judge single hours. Meta says "if you break down the results of your ad set by hour, you may see performance oscillations that average out over time", and suggests "using result breakdowns at the campaign level over the course of a full week." (Understand fluctuations in ad performance).
- Frequency. Meta warns: "If performance begins to drop as your frequency numbers rise, your target audience may be experiencing ad fatigue" (Frequency).
- Fatigue statuses. Meta's Delivery column shows Creative limited or Creative fatigue when cost per result rises against your past ads (About creative fatigue).
Reported by an advertiser on 16 June 2022: about 206 sessions in 30 days averaging about 20 seconds, visits "always on the hour" and no sales, mainly from a Facebook campaign (thread). The numbers are the poster's. Held for a week, a pattern like that is a candidate for Step 5 once Step 4 is clean; it is not proof.
If hours look normal and frequency is low, go to Step 4.
Step 4: Where in the funnel do shoppers drop?
Read your funnel events in order, because where the drop happens separates shoppers who looked and left from shoppers who tried to pay and stopped. This is the check for link clicks but no add to cart.
Meta's standard events, in funnel order (Meta Pixel reference):
- ViewContent: "A visit to a web page you care about (for example, a product page or landing page)."
- AddToCart: "When a product is added to the shopping cart."
- InitiateCheckout: "When a person enters the checkout flow prior to completing the checkout flow."
- Purchase: "When a purchase is made or checkout flow is completed."
The drop map, in our reading:
- ViewContent, almost no AddToCart: the offer, price, page or a buried buy button.
- AddToCart, no InitiateCheckout: usually the cart, shipping cost or trust. A sudden spike in one placement or device can also be automated add-to-carts (see the fake pixel events guide in Step 0).
- InitiateCheckout, no Purchase: checkout friction or payment.
No Meta page we know says how bots move through these events.
When Facebook ads are not converting, give each cause one check:
- Offer and price. On 27 April 2025, an advertiser with little to no purchases feared buyers could find the same item on a large marketplace (thread). Check: compare your price where buyers would shop.
- Page speed and checkout friction. Check: on a phone inside the Facebook app, time the buy button, then place a test order and note every extra field, login wall or failed payment option.
- Learning phase. Meta says ad sets in learning "are less stable and usually have a higher CPA" (About the learning phase for ad sets). Check: the Delivery status.
- Retargeting and prospecting. Our advice, not Meta's: a blend can hide a prospecting-only problem. Check: compare purchases for website-visitor audiences against broad audiences.
A traced example. On 9 March 2026, an advertiser reported about $70 spend, a 2.30% CTR, 5 add to carts, frequency 1.11 and no sales (thread). The numbers are the poster's.
Through the tree: the poster's session recordings mean pages loaded (Step 1) and showed visitors struggling to add to cart; a reply found the buy section far down the page (Step 4). Step 5 never opens.
If the drop has a page, offer or checkout cause, fix that first. If it does not, and the shortfall sits in one placement, device or hour band, go to Step 5.
Step 5: When should you suspect invalid traffic?
Open this branch only after Steps 0 to 4, and look for click-level evidence, not another ratio.
Each row sets a symptom against both readings: funnel or measurement, and invalid traffic.
| Symptom pattern | More likely funnel or measurement when | More likely invalid traffic when | The check |
|---|---|---|---|
| High CTR, no purchases | The shortfall is spread across placements and devices | One placement or device holds the clicks and none of the purchases | Placement and Impression device breakdown; purchases per row |
| Page views, almost no AddToCart | Visitors scroll and leave: offer, price or buy button | Sessions are near zero seconds with no scroll, in one placement or device | The Step 4 drop map; session behavior by placement |
| AddToCart or InitiateCheckout, no Purchase | Shoppers reach checkout and stop: shipping, payment, trust | Click-level evidence shows bots reaching checkout | Checkout test on a phone inside the Facebook app |
| Clicks bunch at the same hours every day | The pattern matches your buyers' normal day | Visits land on the hour or while your buyers sleep, for a full week | Time of day (viewer's time zone) on clicks; store orders by hour |
| Results fall as frequency rises | Delivery shows Creative limited or Creative fatigue | Not a fraud signal: rising frequency points to fatigue | Frequency column; delivery status |
Sources: the Meta pages above; the "more likely" columns are our reading.
Our switch rule, a rule of thumb and not Meta's: move here when the same placement, device or hour band keeps clicks with no purchases for at least a full week, after tracking, objective, page loads and the funnel check out. There is no numeric threshold: we found no Meta page that publishes a normal ratio between clicks and purchases, and we do not publish one either.
Page-side signs reported on the forum: "80+ link clicks" from a Meta Sales campaign, a "100% bounce rate" and near-zero session time (22 June 2024, thread). A 0-second session can also mean the analytics tool recorded no engagement: a sign to check, not proof that bots are clicking your Facebook ads.
Meta's invalid clicks page counts repetitive or accidental clicks, and clicks from fake accounts, bots and scrapers, as invalid. For credits, see whether Meta refunds invalid clicks; for the term, see invalid traffic.
What ClickFortify shows on this branch, and what it cannot
ClickFortify recognizes Meta ad clicks by their fbclid and scores each visit that loads your landing page on 100+ device, network and behavior data points. On the click record it flags three suspicious-conversion patterns: a conversion within 2 seconds of the click, a conversion from a visitor already classified as a bot, and a device with 3 or more conversions.
ClickFortify can switch off Audience Network on ad sets with detected fraud (opt-in per ad account), and it pushes flagged publishers to Meta's publisher block list. When the CAPI proxy is enabled, conversions sent through ClickFortify are checked against the visitor's fraud score, and flagged events are dropped or tagged before Meta receives them. The proxy filters the server events sent through it; pixel-fired browser events still reach Meta.
ClickFortify does not block IP addresses and builds no audience exclusions on Meta. It cannot see clicks that never load the page or Instant Form submissions, so validate those leads in your CRM. A ClickFortify flag is click-level evidence for this branch; it does not diagnose offer, price, page or checkout problems, which is why Steps 0 to 4 come first. See ClickFortify for Meta Ads.
The verdict on Facebook ads clicks but no sales
Clicks with no sales are a funnel or measurement question first, and an invalid-traffic question only when the shortfall is concentrated and persistent. Work the steps in order, and act on the branch the evidence names.
This guide shows where to look, not how often each branch is the answer. Before Black Friday, pair this with the Black Friday click fraud checklist. For every related guide, see all Meta Ads fraud guides.




