How do I set up conversion tracking correctly?

1) Start with a conversion tracking plan (this is where most “bad data” begins)

Define what a “conversion” is in business terms, not platform terms

Before you install anything, decide what actions truly represent success. In most accounts, the cleanest approach is to choose one primary outcome per funnel stage (for example: a completed purchase for ecommerce, or a qualified lead for lead gen), then track supporting actions (like “add to cart,” “view pricing,” or “form start”) as secondary so you can learn from them without letting them steer bidding.

This matters because automated bidding will optimize toward whatever you tell it is a “primary” conversion. If you accidentally make low-intent actions primary, you’ll get a lot more of them—while actual revenue or qualified leads can stall.

Build a simple conversion architecture using goals + primary/secondary

Conversion actions are grouped into conversion goals, and your campaigns can optimize to goals. Within each goal, set the actions you want bidding to optimize toward as primary, and keep “nice-to-know” actions as secondary. Secondary actions still report (typically in “All conversions”), but they don’t usually influence bidding unless you deliberately use them inside certain goal configurations.

As a rule of thumb: primary should be scarce, meaningful, and hard to fake. Secondary can be plentiful and diagnostic.

Pick the right counting logic (leads ≠ purchases)

Google Ads lets you count conversions as One or Every. This is one of the most common setup mistakes I see in audits.

If you’re tracking leads, “One” is usually right because multiple form submits from the same click can inflate performance and confuse Smart Bidding. If you’re tracking purchases, “Every” is usually right because multiple purchases from the same click are still real revenue.

2) Implement conversion tracking the “correct” way (the method depends on your site + stack)

Choose your measurement source: Google tag, Tag Manager, or Analytics events

Today, the most practical setups typically fall into one of three patterns. The best choice is the one that matches how your site is built and who will maintain it:

Directly with the Google tag: Great when you want a straightforward Google Ads conversion that fires on a thank-you page or a defined event and you can place code on the site (or use a tag-integrated site platform).

With a tag management system: Best when you need clean control over triggers (button clicks, form submits, SPA route changes), environments, and debugging—without repeatedly deploying site code.

By importing Analytics events: Strong choice if your organization already governs measurement through Analytics and you want consistent cross-channel measurement, with the option to create conversions from key events and share them into Google Ads for bidding/reporting.

Use the built-in website scan workflow to avoid missed prerequisites

When you create a website conversion, the platform can scan your domain to detect whether a Google tag is present and whether an Analytics property is available to connect. This workflow helps you avoid the classic problem of building conversion actions first, then realizing the site isn’t actually tagged (or is tagged inconsistently across templates).

In practice, I recommend you treat the scan as a “measurement discovery” step: confirm what’s already installed, decide whether you’ll create conversions from the Google tag directly or reuse Analytics events, and only then finalize the conversion action.

Install tags correctly: sitewide first, then conversion events

For reliable attribution, you need a sitewide implementation—meaning your base tag is present across the site (not only on the thank-you page). Sitewide tagging enables first-party measurement cookies to store the click identifier after an ad click (commonly the GCLID), which is then read when the conversion event fires.

Once sitewide tagging is confirmed, you add the conversion event logic (often called an event snippet or event tag). This is what actually signals that a purchase/lead/signup happened. If you’re tracking a page load conversion, the conversion event belongs on the confirmation/thank-you page. If you’re tracking an interaction (button click, form submit, phone link click), you’ll use an event-based trigger rather than relying on a page URL.

Decide whether to go “codeless” (URL-based) or event-based

URL-based conversions are fast and often perfectly fine when you have a stable thank-you page URL and you don’t need transaction IDs, dynamic revenue, or advanced parameters. Event-based conversions are the better choice when you need accuracy on modern sites (single-page apps, embedded forms, multi-step checkouts) or when you want richer data like revenue, transaction IDs, or lead attributes.

If you’re unsure: ecommerce almost always benefits from event-based tracking because revenue accuracy and de-duplication are too important to leave to URL rules alone.

3) Configure conversion settings that protect data quality (and improve bidding)

Set conversion values the right way (even for lead gen)

Conversion values aren’t just for ecommerce. If you’re a lead gen advertiser, assigning a realistic value to a qualified lead is one of the fastest ways to make optimization more ROI-driven—especially if you’ll graduate into value-based bidding later.

Use a single fixed value if you’re starting out, then move toward variable values when you can pass lead quality tiers or downstream revenue back (more on that below). When you change conversion values, remember that changes generally apply going forward, not retroactively—so document your “value change date” in your reporting notes.

Use conversion windows intentionally (don’t leave defaults on autopilot)

A conversion window is how long after an ad interaction a conversion can still be credited. If your buying cycle is short (for example, a 48-hour promo), shorten the window so reporting reflects reality and optimization isn’t over-crediting late conversions. If your buying cycle is long (high-consideration services), lengthen the window so you don’t undercount.

Most advertisers should also avoid “over-tightening” windows early. If you set the window too short, you can starve Smart Bidding of the conversion signal it needs and make performance look artificially weak. A practical approach is to review your time-lag patterns, then set the window to cover the bulk of converting behavior.

Prevent duplicate purchase conversions with transaction IDs

If you track purchases, duplicate conversions are a silent killer—especially when users reload confirmation pages, revisit order history pages, or your site fires the purchase event multiple times.

Use a transaction ID (sometimes referred to as an order ID in certain offline workflows) in your purchase conversion so duplicates can be deduplicated. The key rules are simple: the ID must be unique per transaction, must be dynamically populated (not hardcoded), and must not contain customer-identifying information.

Use conversion value rules only when you have a clear business reason

Conversion value rules let you adjust values based on conditions like device, location, or audience characteristics. This can be powerful (for example, if certain regions have consistently higher average order values), but it’s also easy to overcomplicate. Don’t use value rules to “try to make the account look better.” Use them to reflect true business value differences that your CRM or backend data supports.

4) Turn “basic tracking” into durable tracking (enhanced conversions + offline imports)

Enable enhanced conversions for better measurement resilience

As measurement has become more privacy-constrained, enhanced conversions have become less of an “advanced extra” and more of a best-practice baseline for many advertisers. Enhanced conversions allow hashed first-party customer data (like email) to be used in a privacy-safe way to improve attribution quality.

For ecommerce, enhanced conversions can improve the reliability of conversion measurement. For lead gen, enhanced conversions for leads is especially important when you care about matching leads back to ad interactions and improving bidding accuracy.

If you sell offline (or qualify leads later), import offline conversions the modern way

If a click becomes a lead today but only becomes revenue weeks later in a CRM, you should close the loop by importing offline conversions. This is how you teach Smart Bidding what a qualified lead or a won deal looks like, not just what a form submit looks like.

For lead-based offline measurement, the current best practice is to upgrade toward enhanced conversions for leads workflows, where you can use click IDs and hashed user-provided data to improve match rates and measurement durability. This usually requires three things: capturing the click identifier, capturing the user-provided data at the time of lead (like email), and then importing the downstream outcome from your CRM or file-based source.

Privacy and consent: make sure tags can legally and technically fire

Conversion tracking must respect user consent and privacy requirements. If your consent signals are missing or misconfigured, you can see reduced measurement, limited remarketing eligibility, or missing conversion recording depending on the exact setup. The practical takeaway is that your consent implementation can directly affect whether conversions are recorded and whether they’re eligible for personalization features.

If you operate across regions (even if you’re based in the United States), align your legal/compliance team and your technical implementation so measurement, remarketing, and attribution work as intended for each user segment.

5) Validate, debug, and maintain accuracy (the part most guides skip)

Do a controlled test conversion and verify the right conversion action fires once

Your first test should answer three questions: did the conversion fire, did it fire once, and did it fire on the right action (not a different conversion action with a similar name). This sounds basic, but it’s where most misattribution starts—especially when accounts have legacy conversions, duplicated tags, or multiple containers.

• Confirm tracking status in the conversions table and investigate any “unverified” or “tag inactive” states.
• Use the built-in Tag Assistant workflow from within the account to connect to your site, trigger the conversion, and see real-time debug output.
• Check for double-fires caused by thank-you page reloads, duplicated triggers, multiple tags firing the same action, or SPA route-change behavior.

After launch: watch for common “silent breaks”

Even correct setups can degrade over time. The usual culprits are website redesigns (new templates missing the base tag), form provider changes (embedded forms that bypass your triggers), checkout changes (purchase events fired earlier than confirmation), and consent banner updates (tags running in restricted mode unexpectedly).

I recommend a lightweight monthly cadence: spot-check a live conversion path, review tracking status, scan for sudden shifts in conversion rate by device/browser, and confirm that your primary conversions still match what the business considers success.

Quick reality check: if you want ROI, don’t optimize to “easy” conversions

Conversion tracking isn’t just a reporting feature—it’s the steering wheel for Smart Bidding. The “correct” setup is the one where primary conversions represent true business value, are deduplicated, have sensible counting rules, use appropriate windows, and are validated end-to-end. When you get those fundamentals right, everything else in Google Ads—bidding, budget allocation, and performance analysis—gets easier and more profitable.

Once you’ve defined what a “real” conversion is for your business, chosen the right counting method (for example, “One” for leads and “Every” for purchases), implemented tags cleanly (sitewide base tag plus an event on the true conversion step), assigned values and windows that match your buying cycle, and validated that each action fires exactly once, the ongoing challenge is keeping everything accurate as your site, consent setup, or funnel evolves; that’s where Blobr can fit naturally into your workflow by connecting to your Google Ads account and continuously checking performance and setup patterns, then surfacing clear, prioritized fixes and opportunities through specialized AI agents (including agents that help align keywords, ads, and landing pages), so your tracking and optimizations stay tied to the outcomes that actually matter.