Google Ads, Analytics or My CRM: Which Lead Count Should I Report?
Your monthly report is due, and Google Ads, GA4 and your CRM each show a different lead total for the same campaign. Pick one and someone produces a dashboard that disagrees. Try to reconcile all three to the last lead and you lose a day chasing a gap that never fully closes.
Report the CRM count as your lead number of record, ideally split into qualified leads, because it is the only system that counts people rather than events, filters out junk and sees calls and offline enquiries. Use Google Ads conversions for bidding and GA4 for cross-channel journeys, and state the expected gap instead of forcing a match.
What does each system actually count?
The three numbers disagree because each system answers a different question. The full list of reasons the platforms diverge is covered in our guide to why Google Ads and Google Analytics conversions don't match. For choosing a reporting figure, what matters is what each one is built to count.
Google Ads counts events it credits to ads
Google Ads counts conversion events and credits them to its own ads. It has no view of any other channel, so if a click preceded the form fill, Google Ads claims the lead. Depending on how the conversion action is set to count, one prospect who fills in a form, calls and then sends a chat message can register as three conversions.
Google Ads also dates a conversion to the ad interaction rather than the submission. A lead submitted late in the month can therefore land in an earlier reporting period. The platform models conversions it cannot observe because of consent refusals or browser protections. Native tracking can also include view-through conversions from people who saw an ad but never clicked it.
Spam form fills count too, even when your CRM quarantines them. Taken together, these behaviours push the platform figure higher than the CRM total.
GA4 counts sessions it was allowed to see
GA4 works from anonymous sessions and events. It loses visitors who refuse cookies or browse with tracking blocked. It also spreads credit across every channel, so paid search usually gets less credit than Google Ads gives itself. Phone calls, WhatsApp messages and sales-team enquiries never appear unless you send them in.
If you need the mechanics of why those two Google figures move apart, see why Google Ads and GA4 report different conversion numbers.
Your CRM counts people
A CRM is identity-first. It records a named contact, merges duplicates, lets sales mark junk and logs calls and offline enquiries next to form fills. Its weakness is not the count but the source.
A native "Web" lead source lumps paid search, paid social, organic and email into a single bucket. If the click ID was never captured on the form, the lead exists in the CRM, but nothing ties it back to the campaign that produced it.
Why the CRM should be your number of record
The lead count in a report answers one question: how many potential customers did marketing produce? Only the CRM can answer that in units the rest of the business recognises. Sales works from it, finance reconciles revenue against it, and a lead in the CRM is a person someone can phone.
A platform figure fails the simplest audit. When the sales director asks for the names behind the number, an ad platform cannot produce them, because some of its conversions are duplicates, spam, modelled estimates or view-throughs.
There is also a structural reason. Each ad platform grades its own homework. Google and Meta will both claim a lead that touched both, so summing platform totals double counts the same people. The CRM is the one place where a lead is recorded once, whoever claims it.
Where you can, report qualified leads rather than raw form fills. Join CRM stages to campaigns for the first time and a familiar pattern tends to appear: the campaign with the cheapest cost per lead turns out to have the weakest lead quality.
A funnel reported from the CRM shows the real shape of the pipeline:
- leads
- marketing-qualified leads
- sales-qualified leads
- won deals
That shape is invisible in any platform count.
When is the platform number the right one?
Which conversion number to trust depends on the decision you are making. The CRM wins for reporting. It does not win everywhere.
Bidding and pacing. Smart Bidding optimises on whatever sits in Google Ads, and it learns best from signals that arrive close to the click. CRM outcomes lag by days or weeks, so day-to-day budget and bid calls rightly use platform data.
Comparisons inside one platform. Comparing campaign A with campaign B inside Google Ads is fair even though the totals are inflated. The bias applies to both sides of the comparison.
A CRM you cannot vouch for. The CRM is only the number of record if you can trust four things:
- every form, chat and booking tool writes to it
- duplicate rules exist
- sales actually logs inbound calls
- source data is captured when the lead is created
If any of these fail, the CRM undercounts. In that case, fix the plumbing before you change what you report. A clean platform figure clearly caveated is better than a CRM count that silently misses a third of your enquiries.
Long sales cycles. Google Ads conversion windows are capped. A B2B deal that closes after the window never flows back to the platform. That is one more reason revenue reporting has to come from the CRM, even when the platform drives the bidding.
How to report all three without a fight
Pick the number of record once, define it in the report header and stop relitigating it every month. Then make the other two numbers useful rather than competing:
- Capture click IDs on every lead type. Store the GCLID, GBRAID or WBRAID, along with UTM parameters, on the CRM record the moment a lead is created. Do this for forms, chats, texts and tracked calls alike. That lets the CRM attribute leads to Google Ads itself instead of relying on a generic source picklist.
- Standardise source values. A two-level lead source field works well: a short master list such as paid search, paid social and events, with a detail field for the specific campaign. It keeps reports readable without losing the drill-down.
- Send CRM stages back to Google Ads. Import qualified leads and won deals as offline conversions, with values that reflect what each stage is worth. Bidding then starts optimising toward the same definition you report on, and the two numbers converge in meaning even if they never match in volume.
- Track the gap as a ratio, not a reconciliation. Record the ratio of platform conversions to CRM leads each month. A stable ratio is normal. A sudden jump is a signal worth investigating: a tag broken by a site update, a spam wave, or a new form that never connected to the CRM.
- Never add platform totals together. Report each platform's figure as that platform's claim, and the CRM figure as the total.
The goal is not one number everywhere. It is one number of record, two diagnostic figures and a documented reason for the difference between them.
Conclusion
Google Ads, GA4 and your CRM disagree because they count different things: ad-credited events, observable sessions and identified people. Report the CRM figure, preferably as qualified leads. Let Google Ads drive bidding with CRM stages fed back to it, use GA4 for journey context, and watch the ratio between them instead of trying to make them equal. A report built this way survives the question "who are these leads?", which is the one question that ends most arguments about lead counts.
See how Companion answers lead questions from Google Ads, Analytics and HubSpot data in one grounded view, so each report starts from the same number of record.
Frequently Asked Questions
There is no reliable industry benchmark, because the gap depends on your counting settings, consent rates, spam levels and how many leads arrive by phone. The useful benchmark is your own history. Record the ratio of Google Ads conversions to CRM leads for three months after your tracking is clean. Treat that band as normal and investigate any month that falls outside it.
Rarely, but it can be the best available option when there is no CRM at all, such as a small business whose form submissions go straight to an inbox. GA4 at least applies one attribution view across every channel. Label it clearly as a session-based estimate that misses phone calls, offline enquiries and visitors who refused tracking, and move to a CRM count once one exists.
They can if they are your only signal. CRM outcomes such as qualified leads or signed contracts often arrive days or weeks after the click, and bidding algorithms learn best from conversions close to it. The common approach is to keep a near-click conversion, such as the form fill, in the bidding mix. Imported later stages then carry higher values that shift optimisation toward quality.
Get a complimentary audit to uncover AI opportunities hidden in your data.
Put these strategies to work