Google Ads Conversions and All Conversions: Full Guide for Analysis
Table of contents
- Understanding Conversions and All Conversions
- Ecommerce: Purchase as Primary, Everything Else as Signal
- Lead Generation: Optimizing for the Middle When You Actually Care About the End
- Conversion Actions: Working with Granular Data in the Interface and API
- Conversion Rate and Cost Metrics
- Conversion Value: Making the Numbers Mean Something
- Conversion Value Metrics: Rates and Cost Ratios
- Metric Availability by Campaign Type
- Practical Takeaways
Understanding Conversions and All Conversions
Every Google Ads account tracks goals. A purchase completes. A lead form submits. A free trial begins. Every one of those events fires a conversion tag and gets recorded somewhere in the system. But not all of them end up in the same column, and that difference has real consequences for how your campaigns bid, how your performance looks on paper, and whether you are actually making the right optimization calls.
Google Ads splits conversion reporting into two columns. The Conversions column, or metrics.conversions in the API, contains only the conversion actions you have designated as primary, meaning actions marked Include in Conversions in account settings. These are the goals you consider most business-critical, and they are the only ones the Smart Bidding engine trains on. Target CPA, Target ROAS, Maximize Conversions, and Maximize Conversion Value all optimize exclusively against whatever lands in the Conversions column.
The All Conversions column, or metrics.all_conversions, is the full picture. It includes everything in the Conversions column, plus secondary conversions, view-through conversions, cross-device conversions, and store visit conversions. All Conversions is always equal to or greater than Conversions.
A simple rule: Conversions controls your bidding. All Conversions tells you what is actually happening. You need both to make good decisions.
The gap between the two columns is where most of the interesting analysis lives. A campaign showing 200 Conversions and 840 All Conversions has a 4:1 ratio of secondary or cross-device activity to primary conversions. That ratio tells you something about your funnel, and it matters enormously depending on what business model you are running.
Ecommerce: Purchase as Primary, Everything Else as Signal
In a standard ecommerce setup, the conversion action hierarchy looks something like this:
- Item View (product page view), tracked, secondary, appears in All Conversions only
- Add to Cart, tracked, secondary, appears in All Conversions only
- Begin Checkout, tracked, secondary, appears in All Conversions only
- Purchase, tracked, primary, appears in both Conversions and All Conversions
Only the Purchase event is set as primary. That is the right call. Purchase is the revenue event, it carries a transaction value, and it is the action you want the bidding algorithm to maximize. Setting Add to Cart as a primary conversion action would cause Smart Bidding to optimize for cart additions rather than completed purchases, which typically tanks your ROAS.
But this is where the All Conversions column becomes analytically critical. The ratio between funnel steps tells you where you are losing users.
| Conversion Action | All Conv. | Conv. | Step Conv. Rate |
|---|---|---|---|
| Item View | 12,400 | — | 100% (baseline) |
| Add to Cart | 3,100 | — | 25.0% |
| Begin Checkout | 1,240 | — | 40.0% |
| Purchase | 372 | 372 | 30.0% |
This table shows a 3% overall view-to-purchase conversion rate, which is solid. But it also shows that 75% of users who view a product never add it to the cart, which is the biggest drop in the funnel. That insight does not come from the Conversions column, which only shows 372 purchases. It comes from comparing each step in the All Conversions data broken out by conversion action.
If you only look at the Conversions total, you know your campaign produced 372 purchases. You have no idea whether the funnel is healthy. You cannot tell whether the problem is at the top of the funnel, in the cart experience, or at checkout.
The Conversions column optimizes your bids. The All Conversions column, broken out by conversion action, diagnoses your funnel.
A second critical use of All Conversions in ecommerce is catching upstream pressure. Suppose your purchase volume drops 20% week over week. You check Conversions and confirm that purchases are down. But you segment All Conversions by conversion action and see that Add to Cart is also down 20%, while Item View is flat. That points to a product page issue rather than a checkout issue.
Contrast that with a different scenario where Item View drops 20% while Add to Cart and Purchase ratios hold steady. That points to a traffic quality or impression volume issue, not a site conversion issue. Same Conversions drop, completely different root cause, completely different fix.
Lead Generation: Optimizing for the Middle When You Actually Care About the End
Lead generation campaigns, particularly for SaaS products, have a more complex conversion hierarchy and a more dangerous misalignment risk. A typical funnel might look like this:
- Sign Up Started, secondary
- Sign Up Completed, primary
- Trial Started, primary or secondary depending on the model
- Plan Upgraded / Purchase, primary or tracked offline
The most common configuration is that Sign Up Completed is set as the primary conversion. That is the first tangible commitment a user makes, it is trackable in real time, and it happens frequently enough that Smart Bidding has data to train on. Trial Started may be secondary or primary depending on the product model.
Plan Upgraded, the actual revenue event, is often tracked as an offline conversion uploaded via the Offline Conversions API because the payment happens later in the billing system rather than through a web tag.
This is where the Conversions versus All Conversions distinction becomes a business-critical issue. Your campaign is optimizing for Sign Up Completed. But your business lives and dies by Plan Upgraded. These are not the same event, and the users who complete one do not always complete the other.
| Conversion Action | All Conv. | Conv. | Notes |
|---|---|---|---|
| Sign Up Started | 1,840 | — | Secondary |
| Sign Up Completed | 920 | 920 | Primary, bids optimize here |
| Trial Started | 614 | — | Secondary in this example |
| Plan Upgraded | 107 | — | Offline, secondary or excluded |
The campaign is delivering a CPA of, say, $18 per Sign Up Completed. That looks great. But Plan Upgraded is happening at a rate of 107/920 = 11.6% of signups. So your true cost per paid customer is $18 / 11.6% = approximately $155. Whether $155 is acceptable depends entirely on your LTV, but if your sales team thinks they are buying leads at $18 each, you have a miscommunication problem with real financial consequences.
In SaaS lead generation, the gap between your Conversions CPA and your true revenue CPA is often the most important number in the account, and it only becomes visible when you analyze All Conversions by action.
Important nuance: the long-term solution is to move the primary conversion goal closer to revenue. But you still need enough volume. If Plan Upgraded is too rare for Smart Bidding to learn from directly, the practical compromise is to keep Sign Up Completed as primary while monitoring the upgrade rate as a secondary metric.
The other red flag in lead generation is the Sign Up Started to Sign Up Completed ratio. In the example above, 1,840 users started registration and 920 completed it, which is a 50% completion rate. That is poor for many registration flows. If Sign Up Started is high but Sign Up Completed is low, the registration UX is losing users.
Conversion Actions: Working with Granular Data in the Interface and API
A Conversion Action is the individual goal configuration in Google Ads: the named event, its category, its counting method, its value assignment, and its primary or secondary status. Everything above becomes much more useful when you analyze data at the level of individual conversion actions.
In the Google Ads Interface
In the Google Ads UI, navigate to Goals → Conversions → Summary to see all conversion actions in the account. Each action shows its category, tracking status, whether it is primary or secondary, the conversion window, the attribution model, and whether value is assigned statically or dynamically.
To analyze performance broken out by conversion action in a campaign or ad group report, open the Columns menu, go to the Conversions section, and add the Conv. action segment. This splits each row in the report into separate rows for each conversion action that fired during the selected period. The Conversions column shows only primary actions. The All Conv. column shows everything.
In the Google Ads API
In the API, conversion action data is retrieved by adding segments.conversion_action and segments.conversion_action_name to your GAQL query. When you segment by conversion action, every row in the response represents a specific resource, date, and conversion action combination.
SELECT
campaign.name,
segments.date,
segments.conversion_action_name,
metrics.conversions,
metrics.all_conversions,
metrics.conversions_value,
metrics.cost_per_conversion
FROM campaign
WHERE segments.date DURING LAST_30_DAYS
ORDER BY metrics.all_conversions DESC
A critical API behavior to understand is that when you do not include a conversion action segment, the API returns totals. metrics.conversions becomes the sum of all primary conversion actions. metrics.all_conversions becomes the sum of everything. When you do include segments.conversion_action_name, each row becomes action-specific.
Why this matters: it is common to query totals first and then re-query with conversion action segmentation to diagnose funnel issues. Both views are necessary.
Why You Must Analyze by Conversion Action, Not by Totals
Consider an ecommerce account where someone accidentally set Item View as a primary conversion action. The campaign may show 14,000 Conversions, a CPA of $0.50, and a ROAS of 40x. Those numbers can be mathematically correct while still being business-useless, because most of those conversions are page views rather than purchases.
The same problem happens in lead generation when a Contact Form Opened event is misconfigured as primary. The campaign appears to be producing thousands of conversions at a very low CPA, while revenue stays flat.
Pattern 1: High Item View count with flat or declining Add to Cart
- Product pages are getting traffic but not converting
- Investigate pricing, imagery, reviews, or product descriptions
Pattern 2: Strong Add to Cart with weak Checkout Initiated
- Likely cart abandonment issue
- Investigate cart UX, unexpected costs, or forced account creation
Pattern 3: Strong Sign Up Started with weak Sign Up Completed
- Likely registration flow friction
- Investigate form length, email verification steps, or password requirements
Other patterns matter too. Strong Sign Up Completed with weak Trial Started often points to onboarding problems. Normal trial volume with low Plan Upgraded often points to pricing, value proposition, or product-market fit issues rather than ad delivery problems.
Conversion Rate and Cost Metrics
Google Ads reports a full set of relative and cost-based metrics for both primary Conversions and all-inclusive All Conversions columns. Each has a specific meaning and a specific use case.
Rate Metrics
| UI Name | API Field | Formula | What It Means |
|---|---|---|---|
| Conv. rate | metrics.conversions_from_interactions_rate |
Conv. / Interactions | % of clicks or views that result in a primary conversion. Primary signal for landing page and offer efficiency. |
| All conv. rate | metrics.all_conversions_from_interactions_rate |
All Conv. / Interactions | Same ratio using all conversion actions including secondary. Usually higher than Conv. rate. |
| View-through conv. | metrics.view_through_conversions |
Absolute count | Users who saw, not clicked, a Display or Video ad and later converted. Reflects brand exposure effect. |
| Cross-device conv. | metrics.cross_device_conversions |
Absolute count | Conversions where the click and the conversion happened on different devices. Significant for mobile-heavy traffic. |
Cost Metrics
| UI Name | API Field | Formula | What It Means |
|---|---|---|---|
| Cost / conv. (CPA) | metrics.cost_per_conversion (micros) |
Cost / Conv. | How much you pay per primary conversion. The core efficiency metric for direct response campaigns. |
| Cost / all conv. | metrics.cost_per_all_conversion (micros) |
Cost / All Conv. | Cost per event including secondary conversions. Lower than CPA and useful for upper-funnel cost benchmarking. |
Note on by-conversion-date variants: metrics.conversions and metrics.all_conversions default to reporting on the date of the ad interaction that preceded the conversion. Google also exposes metrics.conversions_by_conversion_date and metrics.all_conversions_by_conversion_date, which report conversions on the date they actually occurred.
Conversion Value: Making the Numbers Mean Something
Conversion value is the monetary amount assigned to a conversion event. It makes ROAS calculable and allows you to differentiate a $20 order from a $2,000 order within the same campaign. Without properly configured conversion values, ROAS numbers are meaningless and Target ROAS bidding cannot function correctly.
Google Ads supports two broad approaches to conversion value: static assignment and dynamic passing.
Static Value Assignment
In conversion action settings, you can assign a fixed value to every instance of that conversion action. This is appropriate for many lead generation workflows where you have a reasonable modeled value per qualified lead. It is not appropriate for ecommerce, because a fixed value cannot distinguish a low-ticket purchase from a high-ticket purchase.
For lead generation, the best practice is to base the static value on historical close rates and average deal sizes.
Dynamic Value Passing
For ecommerce, dynamic value passing is not optional. It is the foundation of accurate ROAS measurement. You must pass the actual transaction value with every purchase conversion event.
Requirement 1: Google tag implementation
- The purchase confirmation page must fire the conversion tag
- For
gtag.js, includevalueandcurrencyparameters in the event object
Requirement 2: Currency consistency and tag scope
- The passed currency must match the conversion action settings
- The tag should fire only on confirmed purchases, not on cart or checkout pages
Requirement 3: Value validation
- Use Google Tag Assistant or Diagnostics to compare recorded values with real order values
- A large mismatch usually means the implementation is wrong
Important nuance: dynamic value passing is one of the most impactful implementation steps for ecommerce accounts. Without it, you are flying blind on ROAS.
For SaaS products where upgrade value is known at the time of payment, the same principle applies. Pass the actual plan price or contract value with the conversion event. If upgrades happen in a billing system outside the website flow, use the Offline Conversions API keyed to the original click’s gclid.
Conversion Value Rules
Google Ads also supports Conversion Value Rules, which allow value adjustments by customer type, device, or location. These rules affect metrics.conversions_value, which means they also affect ROAS calculations and Target ROAS bidding.
Conversion Value Metrics: Rates and Cost Ratios
Once conversion values are configured properly, Google Ads exposes a full set of value-based efficiency metrics that connect campaign performance to business outcomes.
Value Rate Metrics
| UI Name | API Field | Formula | Use Case |
|---|---|---|---|
| Conv. value | metrics.conversions_value |
Sum | Total revenue attributed to primary conversions. |
| All conv. value | metrics.all_conversions_value |
Sum | Revenue across all conversion actions including secondary. Captures full funnel value. |
| Value / conv. (AOV) | metrics.value_per_conversion |
Conv. value / Conv. | Average order or transaction value per primary conversion. Tracks product mix and audience quality shifts. |
| Value / all conv. | metrics.value_per_all_conversion |
All Conv. value / All Conv. | Average value across all conversion types. |
| Conv. value / cost (ROAS) | metrics.conversions_value_per_cost |
Conv. value / Cost | Return on ad spend for primary conversions. Central efficiency metric for ecommerce. |
| All conv. value / cost | metrics.all_conversions_value_per_cost |
All Conv. value / Cost | Full ROAS including secondary attribution. Always greater than or equal to standard ROAS. |
Cost-per-Value Metric
| UI Name | API Field | Formula | Use Case |
|---|---|---|---|
| Cost / conv. value | Computed: cost / conversions_value |
Cost / Conv. value | Inverse of ROAS. Shows how much ad spend it takes to generate $1 of revenue. |
By-conversion-date value variants: metrics.conversions_value_by_conversion_date and metrics.all_conversions_value_by_conversion_date attribute revenue to the date the conversion actually occurred rather than the date of the ad click.
Metric Availability by Campaign Type
The table below shows which conversion and value metrics are available for each campaign type. Partial indicates the metric is available but with restrictions.
| Metric | Search | Display | Video | Shopping | PMax | App | Demand Gen |
|---|---|---|---|---|---|---|---|
| Conversions (primary) | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| All Conversions | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Conv. Rate | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| All Conv. Rate | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Cost / Conv. (CPA) | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Cost / All Conv. | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Conversions Value | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| All Conv. Value | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| Value / Conv. (AOV) | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| ROAS (Conv. value / cost) | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| All Conv. Value / Cost | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
| View-through Conv. | No | Yes | Yes | No | Partial | Yes | Yes |
| Cross-device Conv. | Yes | Yes | Yes | Yes | Yes | No | Yes |
| Segment by Conv. Action | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
A few notes on the Partial and No entries:
- View-through conversions, Search: not applicable because Search ads require a click to serve.
- View-through conversions, Shopping: not supported because Shopping uses click-based attribution only.
- View-through conversions, Performance Max: available for Display and YouTube inventory within PMax, but not isolatable without the channel-level performance report.
- Cross-device conversions, App: app install and in-app action attribution uses a different mobile measurement framework.
Practical Takeaways
Conversions and All Conversions are two views of the same underlying data, separated by the designation of primary versus secondary goals. Neither is more correct. They answer different questions. Conversions tells you what your bidding algorithm is seeing and what it is optimizing toward. All Conversions, broken out by conversion action, tells you what is actually happening across your funnel.
Recommendation 1: Set your primary conversions to the business goal closest to revenue that is still frequent enough for Smart Bidding to learn from.
Recommendation 2: Track meaningful funnel steps as secondary conversions so they appear in All Conversions and help diagnose funnel health.
Recommendation 3: Always segment conversion reports by conversion action before drawing conclusions. A total conversion number without action-level breakdown is incomplete.
Monitor ratios between funnel steps in All Conversions regularly. Changes in those ratios often reveal problems the Conversions column alone will not show. Ensure dynamic value passing is implemented whenever purchase value varies by transaction. And if your true business-critical conversion is too infrequent for Smart Bidding to optimize directly, validate your proxy goal regularly to make sure it remains correlated with revenue.