Why Your Analytics Numbers Contradict Each Other
Evan Betty · · Updated · 7 min read
Google Analytics says you had 40 conversions last week. Your ad platform says it drove 55. Your CRM says you only closed 12 new leads. None of these numbers are necessarily wrong. They're just measuring different things, in different ways, on different clocks.
This is one of the most common things we find in a growth audit: a business making decisions off a number nobody has actually checked. Before you trust any dashboard, it helps to understand why they disagree in the first place.
Here's what actually causes the mismatch, and how to figure out which number to trust for which decision.
Analytics platforms disagree because each one uses its own model, tracking window, and definition of a conversion. Google Analytics, your ad platform, and your CRM are all measuring something slightly different, not all reporting the same event with different levels of error. The fix isn't picking one "correct" tool, it's understanding what each one is actually counting.
In this guide
- Why do different platforms use different attribution windows?
- Are your tracking tags duplicated or missing?
- Why are sessions defined differently across tools?
- How do bot traffic and ad blockers skew your numbers?
- A worked example
- How do you actually check your own setup?
Why do different platforms use different attribution windows?
An attribution window is the amount of time a platform will credit a conversion back to an earlier ad click or visit. Google Ads might credit a conversion up to 30 days after a click. Your CRM might only log the conversion on the day it actually happened, with no window at all. Meta might use its own separate window entirely.
None of these are wrong. They're just answering different questions: "what happened this week" versus "what does this week's traffic eventually produce." Comparing them directly, without accounting for the window, is comparing two different questions and expecting one answer.
Inconsistent link labelling widens the gap further. If your campaign links aren't tagged with UTM tags, GA4 files those visits under "direct" while the ad platform still claims them, so the two tools disagree before attribution windows even come into it.
Are your tracking tags duplicated or missing?
This is the most common technical cause we find, and the easiest to fix once you know to look. A tracking tag installed twice on the same page will double-count events. A tag installed on some pages but not others will silently undercount. Both problems are invisible until someone actually checks tag firing on every page type, not just the homepage.
Google Analytics 4 () tracks activity as events rather than the older pageview-and-session model, which makes duplicate event firing an especially easy mistake to make if a tag manager setup fires the same event through two different triggers. If you're setting up GA4 from scratch, our guide to GA4 for small businesses covers which events actually matter.
Why are sessions defined differently across tools?
A "session" doesn't mean the same thing across every tool. GA4 defines a session differently than the older Universal Analytics did, and your CRM almost certainly defines an active lead or visit differently again. If you're comparing session counts across two dashboards without checking that both use the same definition, you're comparing two different metrics that happen to share a name.
How do bot traffic and ad blockers skew your numbers?
Some of your traffic isn't a person at all: crawlers, uptime monitors, and automated scanners all hit your site and can register as sessions in a tool that isn't filtering them out. On the other side, a real visitor with an ad blocker or strict browser privacy settings may not fire your analytics tag at all, meaning they show up in your server logs or CRM but never in Google Analytics. Both effects push different tools in different directions on the same underlying traffic. We cover how to filter bot traffic out of your analytics separately.
A worked example
Say a 15-person HVAC company runs a lead form on its website. In one week:
| Source | What it reports | Why |
|---|---|---|
| Google Ads | 18 conversions | Counts any form submit within its 30-day click attribution window, including ones from ads clicked the week before. |
| Google Analytics 4 | 14 form-submit events | Counts only submits that fired the tracked event correctly on the site itself, in the reporting week. |
| CRM | 11 new leads | Counts only submissions that reached the CRM as a valid lead record, after removing 3 obvious spam entries. |
None of these three numbers is "wrong." Eighteen is a paid-media attribution figure. Fourteen is a site-behaviour figure. Eleven is a sales-usable figure. The business decision (is this campaign worth the spend) should be made against the CRM number, because that's the one tied to an actual usable lead, not the platform reporting the most generous count.
How do you actually check your own setup?
- Confirm your tracking tag fires once, and only once, on every key page, not just the homepage. Use your tag manager's preview mode to check the pages where conversions actually happen.
- Write down what each platform actually counts (a click, a session, a form submit, a closed lead) and its attribution window, so you're comparing like with like.
- Filter out obvious bot and spam traffic in both your analytics tool and your CRM before comparing numbers.
- Pick one number as your decision-making source of truth for a given decision, usually the one closest to an actual sale or lead, and treat the others as directional context, not competing truths.
If step one turns up nothing at all, rather than just an inconsistent number, that's a different problem. Our guide to why Google Analytics isn't tracking walks through the usual causes.
Key takeaways
- Analytics platforms disagree because they measure different things, not because one of them is broken.
- Duplicate or missing tracking tags are the most common, most fixable cause of a wrong number.
- Attribution windows and session definitions differ by platform, even when the label looks the same.
- For business decisions, trust the number closest to an actual sale or lead, and treat the rest as context.
If you're not sure your own tracking is even set up correctly, that's exactly the kind of thing a growth audit checks first, alongside the other four layers we review. Want a lighter first step? The free growth snapshot gives you a quick read before you commit to anything deeper.
Frequently asked questions
Which analytics number should I actually trust?
The one closest to an actual sale or qualified lead, usually your CRM. Ad platforms and website analytics tools are useful for understanding behaviour and attribution, but the number that should drive spending decisions is the one tied to real revenue.
Why does Google Ads report more conversions than Google Analytics?
Google Ads typically uses a longer attribution window (often up to 30 days) and can credit a conversion to a click that happened well before the reporting period. Google Analytics 4 generally reports only the events that fired within the period you're viewing.
How do I know if my tracking tag is firing twice?
Use your tag manager's preview or debug mode and walk through the actual conversion path (the same page a real customer would use), watching for the same event firing more than once. This is the single most common cause of inflated numbers we see in audits.
Is it normal for my CRM and my website analytics to never match exactly?
Some gap is normal because they measure different things: a session versus a lead record, for example. A small, consistent gap isn't a red flag. A large or unpredictable gap usually means a tracking or spam-filtering problem worth investigating.
Sources
- Google Analytics Help, About GA4 events: https://support.google.com/analytics
- Google Ads Help, About attribution: https://support.google.com/google-ads
