The month is over. The Facebook and Instagram ad report says 50 leads, but your CRM (the program or spreadsheet where managers log customers) shows only 40. The first thought: someone is cheating. In most cases it is not cheating but different counting rules. Real losses happen too, and you can find those yourself.
A word on accuracy. Section names in your account may differ from the ones given here. Where it is the opinion of practitioners (our word for specialists who run ads every day) and where it is the official explanation, we write separately. Where there is no official figure, we say so.
1. Meta counts in its own way
According to Meta’s help, the report uses a default window of “7 days after a click or 1 day after a view”. So a lead is credited to the ad in two cases: the person clicked the ad and left a request within seven days, or only saw the ad without clicking and left a request within a day.
For you this means: Meta credited some leads to the ad although the person could well have come on their own, for example by searching your company name. Your CRM knows nothing about this, it just sees a lead.
There is a second reason that few people talk about. In autumn 2025 practitioners wrote that for Meta a “click” is any action on the ad: a like, a share, a save. But in its announcement of 3 March 2026, Meta says otherwise: for website conversions, a click in attribution will mean only a link click, while shares, saves and other clicks move to a separate category called “engage-through” attribution. According to Meta, the changes roll out in waves during March 2026, so different accounts could get them at different times. The takeaway is simple: if you compare a report for a month before spring 2026, the gap may have been larger than it is now.
2. One person, several leads
A person pressed “Submit” twice. A manager sent a test request to check the form. Someone tapped the wrong thing. The Facebook form pulls contact details from the person’s profile, so they may have submitted an outdated number: practitioners write about autofill, and Meta does not officially confirm it. In Meta’s report that is three leads. In the CRM it is one or none, because the manager does not enter duplicates and tests.
3. The lead arrived but never reached the CRM
This is the only cause where you really lose money. The manager could not get through and did not write it down. Or nobody transfers leads from the Facebook form to the CRM: they sit in the Meta account and are forgotten.
Here is how to check yourself, it takes about twenty minutes:
- Open Meta Business Suite and find the Leads Center section: it is where you manage and download leads.
- Export last month’s leads. There is another route, described by practitioners: in Ads Manager, next to the campaign’s results number, there is a download icon, and leads can be exported as CSV or XLS.
- Compare the phone numbers with the CRM, not the number of rows. The numbers that are not in the CRM are your losses.
One caveat. Some write-ups say Meta keeps form leads for about 90 days and then deletes them. Meta does not officially confirm this, so play it safe: export leads every month, not once in six months.
So that leads do not sit in the account, you can connect a CRM to the form. Meta says leads from ads can flow into your CRM automatically. Practitioners add that you need to respond to a lead fast: one source gives a limit of one hour, and Meta gives no official figure. The later the call-back, the fewer people pick up.
4. The opposite happens too: the CRM has more
A person saw the ad but called two weeks later. Or came from Google or through friends. Meta does not see these leads, and does not have to. There are two technical nuances as well. First, Meta Ads Manager shows only events linked to ads, while the Events Manager section (events your website sends to Meta) shows all of them, including non-ad ones. Second, Ads Manager by default assigns a conversion to the day of the click, not the day of the lead: third-party write-ups say so, and Meta does not officially confirm it. Someone clicked at the end of the month, the lead came at the start of the next, and the two systems logged it in different months.
5. What gap is normal
Nobody guarantees figures here. Meta does not give an official “normal” gap between Meta and a CRM. Practitioners aim at about 20%: these are several independent mentions, but with no dates and no explanation of how it was counted. So this benchmark may be out of date, especially after the March 2026 changes, and it should not be taken as a law.
A rough guide looks like this. If Meta says 50 and the CRM says 40, it looks like different counting rules. If Meta says 50 and the CRM says 15, it is time to look for duplicates, tests, spam and leads that never arrived. In our experience, behind such a gap there is most often not ad fraud but a break between the form and the CRM: leads are not transferred or nobody replies to them. This is our experience, not statistics.
6. What to do yourself
- Match phone numbers, not counts. A count proves nothing, the numbers show which leads got lost.
- Compare the same period in the report and in the CRM, including the last day.
- Look at the Meta report with a different window. In Ads Manager, according to third-party write-ups, there is a mode for comparing attribution settings (Compare Attribution Settings): it shows results with different windows in one report, for example clicks only. The difference between them is the leads that Meta credits to a view. The names of items in your account may differ.
- Turn on SMS number verification in the form. According to practitioners, this is a checkbox “Require SMS verification to submit form”, placed in the “Quality filters” block: without the code from the message the form is not submitted and does not count as a lead. Meta describes this feature in its help (“Enable the SMS Verification Feature to Help Improve Lead Quality”), but in some accounts the option may not be there.
Here practitioners’ opinions differ. Some consider verification the main way to remove accidental and fake leads. Others warn: some real customers will not want to enter a code, there will be fewer leads, and the cost of one lead will rise. Meta gives no official figures on this. So try it on one campaign and compare after a month: how many leads there are, how many of them reached payment. Practitioners recommend the “Higher Intent” form type for the same reason: it adds a confirmation screen before submitting.
Summary
A gap between Meta and the CRM is not a verdict. Part of the difference is Meta’s counting rules, which changed in spring 2026. Part is duplicates and tests. And part is real loss, when a lead sits in the Meta account and never reached the CRM. Export a month of leads, match the phone numbers and connect the CRM to the form so that nothing gets lost.
If you want to see how an ad report is read without a marketer, read the article on what your ads count as a lead. The Meta pixel on your website and other connections are covered in the website checklist, and how long to wait for ad results in the article on timing. To see how we run ads on Facebook and Instagram, go to the Meta advertising section.
Advertising on Google as well? Want a free audit of your Google Ads account? Send us a request: we will check where the money goes and where your ads lose enquiries.