Why Google Ads Conversions Don’t Match Your CRM

Written by: Aaron Rovner, Founder, Saas Hero

Key Takeaways

• Google Ads wrong conversions occur when platform counts diverge from CRM leads because of attribution windows, duplicate firing, view-through credits, missing GCLIDs, or Consent Mode modeled conversions.
• Five mechanisms explain every gap: micro-conversions set as primary, duplicate firing, view-through and attribution bloat, broken GCLID, and Consent Mode v2 modeled conversions.
• A repeatable reconciliation procedure comparing Google Ads (by conversion time), GA4 key events, and CRM records for the same month separates legitimate differences from broken tracking.
• Agencies that stop at the ad account leave conversion tracking unowned, and the five diagnostic questions reveal whether your agency owns the full path from impression to CRM record.
• SaaSHero owns paid media, landing pages, and CRM-connected attribution as one accountable team, which closes the structural gaps that produce wrong conversions.

Talk To A Team That Fixes Wrong Conversions

Introduction And Problem Definition

Google Ads says 100 leads. Your CRM has 35. Your paid media agency says the gap comes from attribution differences. You do not believe it, and you are right to be skeptical, yet you cannot prove which explanation is correct without a structured test.

Google Ads wrong conversions have five main mechanisms: micro-conversions set as primary, duplicate firing, view-through and attribution bloat, broken GCLID, and Consent Mode v2 modeled conversions. This article shows how to prove which one applies to your situation. It ends with an ownership question because the reason this gap keeps coming back is structural: conversion tracking sits outside the scope most paid media agencies are paid to own.

Before reading further, pull one date range, the last complete calendar month, and write down one conversion definition you will use consistently across all three systems.

Get A Conversion Tracking Audit

The Five Causes Of Google Ads Wrong Conversions, Explained By Mechanism

Each cause below produces a wrong number through a specific mechanism. Knowing the mechanism lets you diagnose instead of guess.

Micro-conversions set as primary. In Google Ads, the Conversions column contains only conversion actions designated as primary, those marked “Include in Conversions,” and these are the only goals Smart Bidding trains on. When newsletter signups, content downloads, or low-commitment form fills are set as primary, Smart Bidding finds more of the cheapest people to convert. That group often includes students, job seekers, competitors, and existing customers. A campaign CPA of $18 per signup can translate to a true cost per paid customer of approximately $155 once the actual signup-to-upgrade rate is applied, a gap that only becomes visible when conversion actions are analyzed individually. The platform reports a falling cost per conversion while pipeline stays flat.

Duplicate firing. Double-counting occurs when a native Google Ads conversion tracking tag fires on a thank-you page and the same GA4 purchase event is imported as a separate conversion action, with both set to primary. Every conversion is then counted twice in the Conversions column, reported cost-per-acquisition looks like half of reality, and Smart Bidding optimizes toward a number that does not exist. A Tinuiti internal audit published in May 2026 found that roughly 41% of Google Ads accounts onboarded in the prior 18 months arrived with at least one material conversion tracking error, and 28% had errors significant enough to materially distort Smart Bidding signals.

View-through and attribution bloat. Google Ads’ view-through conversions metric counts users who saw, but did not click, a Display or Video ad and later converted, reflecting brand exposure effect. Google Ads’ data-driven attribution model assigns fractional credit across multiple touches, so the Conversions column shows non-integer values and credits interactions the click did not directly produce. Both behaviors inflate platform-reported numbers relative to CRM-observed outcomes.

Broken GCLID. A gclid is a unique tracking parameter Google Ads automatically appends to a landing page URL when someone clicks an ad; it ties conversions back to the exact ad, keyword, and query that drove them. The most common fault in lead-gen accounts is a GCLID captured on the landing page and lost before the form, because the visitor lands on a blog post, browses to the pricing page, and submits a form that never saw the URL parameter. The conversion is counted in Google Ads but cannot be joined to a CRM record.

Consent Mode v2 modeled conversions. Google Ads presents Consent Mode modeled conversions in the standard conversions column and feeds them into automated bidding, so the conversions column can include estimated rather than observed conversions and will differ from raw observed counts or CRM records. A sudden decline in Google Ads conversions that does not correspond to a change in traffic or campaign performance is the first red flag of a Consent Mode v2 or modeling issue rather than a true demand drop.

Before pulling data, note which of these five your account could plausibly have. Most accounts with a significant gap have more than one.

Find Out Which Cause Applies To You

How To Reconcile Google Ads, GA4, And CRM For The Same Date Range

This reconciliation procedure is the diagnostic step no competitor teaches. Run it before drawing any conclusion about your agency or your tracking.

1. Pick one complete calendar month as your date range and confirm the same timezone is set in Google Ads, GA4, and your CRM export. A timezone mismatch can shift day-level numbers without any real discrepancy at the month level.
2. Pull the conversion count from Google Ads using the Conversions (by conversion time) column, not All Conversions. Google describes the Conversions (by conv. time) column as the column to use when comparing with third-party tools. This column gives you the fairest baseline for comparison.
3. Pull the same event from GA4 as a key event for the identical date range. In GA4, a key event is counted across every traffic source that touches the property, while a Google Ads conversion is counted only when Google Ads’ own attribution model credits one of its ads with the outcome. Expect the GA4 count to sit above the Google Ads count.
4. Pull the corresponding leads or opportunities from your CRM for the same date range, filtered to the same conversion definition, such as “New Lead Created” or “Demo Requested.” This number represents real people after deduplication and validation.
5. Compare system by system and record the gap at each hop: Google Ads vs. GA4, then GA4 vs. CRM. Do not compare Google Ads directly to the CRM without the intermediate step, because you will miss where the gap actually appears.

A Google Ads vs. GA4 gap points to tag configuration issues, view-through attribution, or Consent Mode modeling. A GA4 vs. CRM gap points to form-to-CRM handoff failures or GCLID loss. If GA4 and ad platforms agree but the CRM is lower, investigate backend writes, lead validation rules, or duplicate suppression. If the CRM is higher than analytics, users may be converting through paths the front-end tracking never captured.

Write the three numbers and the two gaps on one line before drawing any conclusion.

How To Tell A Legitimate Attribution Difference From Broken Tracking

Attribution differences form a real category. The same phrase also gets used to dismiss a marketing leader when the tracking is genuinely broken. The test below settles the argument.

Legitimate attribution difference: The discrepancy is stable month over month within a predictable range, explainable by a known window or model difference, and the CRM still receives every lead. No records are missing, only the credit assignment differs. A healthy conversion gap is stable and explainable. The platforms sit above the CRM by roughly the same proportion each month, and the direction never flips. Most unmatched rows classify as window and timestamp effects.

Broken tracking: Leads are missing from the CRM entirely, the same conversion fires twice, or the gap widens over time without a corresponding change in traffic or campaign structure. A broken conversion gap shows a step change that jumps on a particular date rather than drifting, indicating a deployment, a consent banner change, or a tag that stopped firing, which has a cause findable in a changelog.

A stable gap of 10–20% that is consistent month over month is generally attributable to structural differences. A gap of 2x or 3x, or one that appeared suddenly, is almost always double-counting or a broken link.

Classify the gap as stable-and-explainable or widening-and-missing before escalating to your agency or your board.

Have SaaSHero Review Your Attribution Gap

How To Fix Destination Mismatch In Google Ads And Remove The Wrong Conversions

Destination mismatch occurs when the conversion action name, the tag trigger, or the destination URL does not match what you intended. The procedure below also covers how to remove conversions that inflate your primary count.

1. Open conversion actions in Google Ads under Goals → Conversions → Summary. In the Google Ads UI, conversion actions are viewed under Goals → Conversions → Summary, and performance can be broken out by conversion action by adding the “Conv. action” segment from the Columns menu.
2. Identify which actions are currently set as primary. List every primary conversion action and confirm it maps to a qualified pipeline event, such as a demo request or opportunity created.
3. Demote low-commitment actions to secondary status. Secondary conversions are tracked and visible in reporting but excluded from account-wide Smart Bidding optimization. Google’s primary and secondary conversion actions page states that primary actions are reported in the Conversions column and used for bidding, while secondary actions appear only in All Conversions and are not used for bidding.
4. Keep only the events that map to qualified pipeline as primary. For most B2B companies, that means a demo request, a sales-qualified lead stage in your CRM, or an offline conversion import tied to a meaningful lifecycle event.
5. Verify the destination URL and conversion action name match exactly, and confirm the tag fires once, not on page load and not on every refresh, using Google Tag Assistant.

Demote first, delete later. Secondary status preserves the historical data while removing it from bidding. Deleting a conversion action erases its history and can disrupt Smart Bidding’s learning period.

What To Ask Your Paid Media Agency When Conversions Do Not Match Your CRM

Five questions distinguish an agency that owns the path from impression to CRM record from one that stops at the ad account.

1. What conversion actions are set as primary? A good answer names them and explains why each maps to qualified pipeline. A deflection sounds like: “Google handles that automatically.”
2. Are lifecycle-stage events pushed back into Google Ads? A good answer describes the CRM-to-platform connection and how a lead becoming a sales-qualified lead triggers an offline conversion import. A deflection sounds like: “We optimize to the platform’s conversions.”
3. Who owns the landing page and the form? A good answer is: “We do.” A deflection is: “That is your web team.” Agencies that do not own landing pages are almost impossible to hold accountable for conversion performance.
4. How do you reconcile Google Ads to the CRM? A good answer describes a monthly reconciliation using the procedure above. A deflection is: “Attribution differences are normal.”
5. What does the monthly report lead with? A good answer leads with pipeline and cost per qualified lead. A deflection leads with leads, CPL, and impression share.

Ask all five in one call and write down which ones receive deflections. The pattern of deflections tells you where the scope boundary sits, and that boundary is usually drawn by the retainer, not by the people executing the work.

The table below contrasts what an agency optimizing to form fills reports versus one optimizing to CRM data, so you can see which pattern your agency matches.

Talk To A Paid Media Team That Owns The Full Path

Why This Keeps Happening: The Scope And Fee Structure That Leaves Conversion Tracking Unowned

Wrong conversions are primarily an ownership problem, not a tracking problem. The conventional paid media agency retainer is scoped to the ad account. The landing page belongs to the client, the CRM to RevOps, and the conversion definitions to whoever configured tag manager, often years earlier and often no longer at the company. Everyone executes their scope faithfully and still produces a result nobody is accountable for.

A paid media management fee should explicitly cover named channels, campaign setup, optimization, testing, reporting, meetings, and strategic oversight. Creative production, landing pages, and tracking implementation should be identified in the proposal as either included or separately billed. In most agency contracts, tracking implementation is not included. It degrades silently, and nobody is positioned to notice because nobody owns the chain.

Agency scope typically covers driving traffic, not converting it: conversion rate optimization, landing page builds, and advanced attribution modeling are often excluded from standard retainers. The boundary is also held in place by how the work is priced. Percentage-of-spend models create a structural conflict of interest by rewarding budget expansion over media efficiency. Per-channel pricing means testing a new channel raises the client’s fees before it has returned anything, so budget calcifies where it was first placed.

Agency Optimizing To Form Fills Vs. CRM Data

An agency optimizing to form fills reports falling cost per lead while pipeline stays flat. The ad platform behaves correctly for the goal it was given. A Google Ads account that optimises on form fills is bidding for enquiries, not customers. An agency optimizing to CRM data reports pipeline, CAC, and payback period, the numbers a board actually asks about.

SaaSHero is the outsourced inbound growth team for B2B companies that owns the entire path from impression to CRM record: paid media, creative, landing pages, and CRM-connected attribution all sit under one accountable team. As a Google Premier Partner (top 3% of agencies) and G2 High Performer in digital marketing for over two years, ranked #20 of approximately 6,000 agencies, SaaSHero has served 100+ B2B companies and managed over $60M in lifetime ad spend.

SaaSHero separates primary from secondary conversions and uses only primary conversions for account-wide optimization. That distinction matters because it determines what Smart Bidding actually learns from. To close the loop, SaaSHero also pushes lifecycle-stage events back into the ad platforms, so bidding trains on qualified outcomes rather than form fills. Because the post-click experience is part of the same scope, SaaSHero owns landing page design, copy, build, hosting, and testing, so no third-party backlog can stall the path from click to conversion. Reporting runs in the client’s own CRM (HubSpot, Salesforce) with Looker Studio dashboards showing pipeline, CAC, and payback period, so the metrics that matter to the board are visible without a separate reconciliation exercise. Finally, the retainer is indexed to total monthly ad spend rather than channel count, which means reallocating budget carries no fee consequence and the incentive to test new channels is not penalized. Every engagement includes CRM-connected attribution reporting built to answer the questions a board actually asks.

Map who owns each hop from impression to CRM record on your current engagement. If any hop has no owner, that is where the gap lives.

Schedule A Pipeline-Focused Strategy Call

When To Fix The Paid Media Agency Relationship Versus When To Leave

If the agency owns the tracking, the landing page, and the CRM connection, and the gap is stable and explainable, the relationship is worth fixing. Work through the reconciliation procedure together, demote the wrong primary conversions, and establish a monthly reconciliation cadence.

If the agency’s scope stops at the ad account and it cannot answer the five questions above, the structure, not the people, is the problem. Likable people executing the wrong scope will produce the same gap next quarter. Your decision is about structure and accountability.

Get Help Rescoping Or Replacing Your Agency

Before You Escalate: A Checklist

• One date range selected and documented
• One conversion definition agreed across all three systems
• One timezone confirmed in Google Ads, GA4, and CRM export
• Conversions (by conversion time) column used in Google Ads, not All Conversions
• Primary vs. secondary conversion audit completed in Goals → Conversions → Summary
• GCLID presence confirmed in CRM records for a sample of recent leads

Conclusion And Next Steps

The reconciliation procedure above proves the discrepancy. The five-cause framework identifies the mechanism. The five agency questions expose whether your paid media agency is accountable for the fix. If the gap comes from broken tracking or agency configuration, the structural fix is one team owning the path from impression to CRM record, not a tag audit handed back to whoever configured it last.

See How SaaSHero Connects Your CRM To Google Ads

Frequently Asked Questions

Why Does Google Ads Always Report More Conversions Than My CRM?

Google Ads and your CRM measure different things. Google Ads counts every conversion event it can attribute to an ad click within its attribution window, including view-through conversions, cross-device conversions, and modeled conversions estimated for users who declined consent. Your CRM counts unique people who became real leads after deduplication, spam filtering, and lead validation. A prospect who clicks a Google ad and a LinkedIn ad in the same week can appear in Google Ads’ conversion count while appearing once in your CRM. Modeled conversions add another layer: Google estimates conversions it could not directly observe and reports them in the same column as observed conversions, with no visual distinction. The gap between Google Ads and your CRM is therefore partly expected and partly diagnostic. A stable gap of 10–20% that is consistent month over month is generally attributable to these structural differences. A gap of 2x or 3x, or one that appeared suddenly after a tag, container, or attribution-setting change, is almost always double-counting or a broken link rather than normal platform-behavior differences.

What Is The Difference Between Primary And Secondary Conversion Actions In Google Ads?

As covered in the Five Causes section, primary conversion actions are the only goals Smart Bidding trains on. The practical consequence is that if a newsletter signup is set as primary, Smart Bidding will find more people likely to sign up, not more people likely to become qualified pipeline. Secondary conversion actions still appear in reporting but do not influence bidding. For most B2B companies, the correct primary set is a small number of high-intent events, while everything else should sit as secondary.

What Is GCLID And Why Does It Matter For CRM Attribution?

GCLID stands for Google Click Identifier. It is a unique parameter Google Ads automatically appends to a destination URL when someone clicks an ad. Capturing the GCLID on the lead form and storing it as a custom field in your CRM makes it possible to join a CRM record back to the specific ad, keyword, and campaign that produced it. Without the GCLID in the CRM, you can see that a lead exists but you cannot prove which ad drove it. The GCLID is also the key that unlocks offline conversion imports. When a lead reaches a meaningful stage in your CRM, such as qualified, opportunity created, or closed-won, you can upload the GCLID back to Google Ads as an offline conversion, which teaches Smart Bidding to find more people like your actual customers rather than more people who fill out forms. Common failure modes include redirects stripping the query parameter, cross-domain forms that never see the original URL, and CRM workflows that lowercase or trim the value (the GCLID is case-sensitive). A GCLID debugging checklist should confirm auto-tagging is enabled, the hidden form field is populated, no redirect strips the parameter, and the conversion window covers the elapsed time between click and upload.

How Does Consent Mode v2 Affect My Google Ads Conversion Count?

Consent Mode v2 is Google’s framework for handling conversion measurement when a user declines cookie consent. When a user declines, Google cannot set advertising cookies or read existing ones. In advanced Consent Mode, the Google tag still sends cookieless pings, signals that an interaction occurred without identifying the individual, which Google uses to model estimated conversions for the non-consenting population. These modeled conversions appear in the standard Conversions column alongside observed conversions, with no visual distinction. The result is that your Google Ads conversion count includes a mix of directly observed events and statistical estimates, while your CRM contains only real people who submitted a form. In markets with high consent refusal rates, common in Germany, France, and the Netherlands, the modeled share of reported conversions can be substantial. A common implementation failure is setting the default consent state too late, after Google tags have already started collecting data, causing Google to treat the user as having no consent signals at all. If your Google Ads conversions dropped sharply on a specific date with no corresponding change in traffic or campaign structure, a Consent Mode v2 misconfiguration is the most likely cause.

Should My Paid Media Agency Own Conversion Tracking Setup And Maintenance?

Conversion tracking setup and maintenance should be owned by whoever is accountable for the performance of the campaigns that depend on it. In practice, the conventional paid media retainer is typically scoped to the ad account, with tracking implementation, landing page development, and CRM integration owned by other parties rather than the agency. This structure creates a gap: the agency is judged on conversion performance but does not own the infrastructure that produces the conversion signal. When tracking degrades, because a site redesign removed a tag, a redirect started stripping the GCLID, or a Consent Mode update was misconfigured, the agency has no contractual obligation to catch or fix it. The result is that conversion tracking errors accumulate silently, Smart Bidding trains on increasingly corrupted data, and the gap between platform-reported conversions and CRM records widens. The correct structure is one team owning paid media, landing pages, and CRM-connected attribution together, so the path from impression to CRM record has a single accountable owner. If your current agency cannot answer what conversion actions are set as primary, who owns the landing page, and how they reconcile Google Ads to your CRM monthly, the scope is the problem.