6 Steps to Set Up Conversion Tracking in Google Ads

Clicks can feel great, but they don’t pay the bills. You might have traffic pouring in and the bank account staying flat because you’re not measuring the actions that matter. This short guide gets straight to it: six clear steps to turn clicks into measurable results.

Built for 2026 realities, we cover automation, privacy changes, consent mode and fewer freebies from cookies. I’ll show how to pick the right outcomes, install a tag the proper way and test like you don’t trust anyone.

If you want a hand, check /google-ads-management-adelaide for my services in Adelaide. Later on I’ll include a real client result so you know this isn’t just theory, and a comparison table to help you choose between a manual install and using google tag manager.

• Pick your conversion types first.
• Install the tag properly.
• Set clear values for each action.
• Choose an attribution model that suits your goals.
• Test thoroughly and don’t trust defaults.

Not every action on your site is worth counting — let’s clear up what really matters.

Definition: A conversion is any user action that directly advances your business goal, like a sale or a qualified lead.

What doesn’t count: time-on-site, pageviews that aren’t goal-related, social shares and broad video views — these can be useful signals, but they’re vanity unless they lead to real outcomes.

Micro actions show intent: a pricing page view or newsletter signup. Macro actions close the deal: a tradie quote request or an online purchase.

For a small bakery, a macro could be an online order; a micro might be a visit to the menu page. For a plumber, a macro is a booked job via a lead form, while a micro is a call button click.

• Lead form submission (lead form)
• Phone call from ads or website
• Purchase or subscription confirmation
• Live chat start, thank-you page view, or specific button clicks

Conversion categories are mainly a sorting and goal-setting system — they don’t do magic. Pick Sales for purchases and Leads for enquiries so the platform knows what to value.

Rule of thumb: if you’d happily pay for that action, track it as primary. Track too many actions and you inflate your report numbers — which makes cost-per-lead look better than reality.

Next: when you set proper priorities, the ad system spends differently to find the outcomes you care about.

Give the system outcomes, not just traffic, and it will hunt the people who pay. Without outcome data the platform simply chases clicks. With proper conversion tracking, budgets move toward leads and sales you actually value.

Once set up you can diagnose which ads, keywords and campaigns deliver real results — not just noisy visits. That clarity makes budget shifts obvious, and you make better calls about where to spend.

Smart bidding automation needs a clear goal. If the goal is wrong, the machine will confidently optimise for the wrong thing at speed. Give it the right outcome and those bidding strategies start to work for you.

Keep ROI honest by assigning values and marking primary vs secondary actions. Track too many soft events and your cost-per-lead looks falsely low. We’ll show exactly how to avoid that later.

Mini example: an ebook download may look cheap, but it rarely becomes revenue. A phone call over 60 seconds? That could be your highest-value lead.

Want a quick audit of what to fix first? Check /ppc-audit — then you can test changes and watch spend follow the right signals.

Decide what you’ll measure first, then set up the tag to report that cleanly. Pick a small number of actions that match your business model so your reports stay useful, not noisy.

Start here unless you run an app-only business. Track thank-you pages, form submits and key button clicks. These are the default for ecommerce and lead gen — they show up quickly in reports and let bidding systems learn.

Essential for local services. You can count calls that start from an ad click or calls that originate on your site. If most customers call, mark calls as primary so bids reflect real value.

Only add these if you have an app. Track installs, purchases or signup milestones inside the app. Don’t overcomplicate things with app events if your business sells physical goods on a website.

Use this when leads convert later in your CRM. Uploading closed sales ties offline revenue back to clicks and makes bidding smarter — because the system learns which ads produce real customers, not just enquiries.

View-through conversions show influence — someone saw an ad, didn’t click, but converted later. Cross-device conversions credit journeys that start on mobile and finish on desktop. Both stop you under-crediting mobile discovery and display influence.

Quick checklist:

• Lead gen: track form submits, calls and CRM uploads.
• Ecommerce: track purchases and dynamic values; keep “every” count for orders.
• App businesses: track installs and high-value in-app events only.
• Multi-touch sales: include offline uploads and view-throughs to map the full path.

When you choose wisely up front, your reports show real outcomes and avoid inflated numbers. If you need help linking tags and tag manager, follow this guide to link Google Tag Manager to Google.

Before you touch any tags, let’s get your access and settings tidy so setup doesn’t stall.

Why bother? A few minutes of prep saves hours of back-and-forth with devs and avoids publishing changes that don’t work.

• google ads account admin access (or someone who can grant it).
• Website access or a developer who can add code or verify pages.
• Admin rights to google tag manager so you can publish or test.

Make sure tag manager is installed and you know Preview mode basics. Preview lets you see events fire before anything goes live — it stops “hope marketing” from hitting production.

Decide how long after a click you still want to give credit. Windows run from 1 to 90 days; longer sales cycles often need longer windows.

For many Aussie service businesses, don’t default to 7 days unless you’re sure the sale completes that fast.

“Collect access and testing rights up front — it makes creating and verifying actions fast and painless.”

If you’d like help with tag manager setup, see how long reviews take or check /google-tag-manager-help for hands-on support.

Let’s make the actual setup simple: open your account and we’ll add a new conversion action that matches what your business values.

Open the left menu and follow Goals → Conversions → Summary. You’ll see existing conversions and a big button to create new items.

Click that button, then pick the source you need and the setup flow will change to match your choice.

Select Website for on-site thank-you pages, form submits and button clicks. Pick Phone calls if most leads ring you, App for installs, or Offline to upload CRM sales later.

Name things clearly so filters and audits stay sane. Use this pattern: Type | Action | Page. Examples:

• Lead | Contact form | /contact/thank-you
• Sale | Purchase | Checkout success

Good names stop duplicates and make reporting filters simple. You can edit later, but it’s quicker to get this right now than clean months of messy data.

Pick the conversion category that matches intent — Sales for purchases, Leads for enquiries. Then choose an optimisation goal that tells smart bidding what to chase.

Set the counting method and value to reflect business reality, save, and move on to tag installation.

A single, sitewide tag keeps the user journey intact — set that up before you add per-page events. The AW- tag is the bridge between your site and your ads account; it lets clicks be matched to later actions so bidding systems learn what really pays.

Why it must load on every page: attribution and cookies need to be set before a user converts. If the tag only lives on a thank-you page, you break the click-to-sale stitch and reports lie to you.

In tag manager create a new tag, paste the AW- ID, then set the trigger to Initialization — All Pages (or All Pages). Publish and test in Preview mode. This delivers the tag early in the page load so attribution works reliably.

Good news: the conversion linker behaviour is now baked into the main google tag when it fires sitewide. In most setups you don’t need an extra Conversion Linker tag unless you have unusual cookie restrictions.

Quick steps (practical):

• Create the google tag or manual snippet and add the AW- ID.
• If using tag manager, trigger on Initialization or All Pages and publish.
• Don’t add a separate Conversion Linker if the main tag fires sitewide.

Common mistake: installing the tag only on the thank-you page — that breaks click stitching and destroys reliable reports.

Short answer: Create a Google Tag Manager tag using the conversion ID and label, then fire it on the event that proves value — thank-you page for simple sales, a confirmed form submit for leads, or a dataLayer event for the cleanest result.

In GTM, add a Google Ads Conversion Tracking tag and paste the Conversion ID and Conversion Label from your account. The ID is account-level; the label ties to the specific action.

Trigger options:

• Thank-you pageview: simplest for ecommerce receipts or booking pages — reliable if the site redirects to a unique URL.
• Form submit: useful for lead forms but can misfire if the form uses AJAX; confirm with Preview mode.
• dataLayer event: the cleanest for dev help — fires only after the server confirms success.

Set count to One for leads (one person = one lead) and Every for purchases (each sale counts). Example: a customer may request two quotes, but you usually want a single lead counted; a customer can make multiple purchases and each should register.

“Avoid firing on a button click unless you verify the form actually succeeded.”

Anti-bug tip: test in GTM Preview, confirm the tag fires only on the intended event, then publish. That simple process stops noise in your reports and keeps ads conversion tracking useful.

Give your bidding strategies something they can actually work with: numbers that reflect real business value. If you only report that a conversion happened, the system can’t tell which actions make profit.

Use fixed values when leads are fairly uniform — for example, assign $X per qualified enquiry. Use dynamic values for ecommerce where order totals vary. For services with mixed leads, apply tiered or weighted values until CRM data lets you refine them.

Start with a fixed value if you’re unsure, then update once sales data arrives. Real conversion values let Target ROAS and Maximise Conversion Value actually chase revenue instead of cheap clicks.

“One Adelaide service client cut wasted spend by shifting optimisation from micro actions to valued leads.”

When to use “Maximise conversion value”: only once your values are reliable and your data is consistent. If values are guesses, that strategy will optimise for misleading signals. If you want a deeper example, see /case-studies.

If your numbers shift after a settings tweak, don’t panic — attribution often explains it. Change the model and the same user journey can credit different touchpoints, so reports will look different overnight.

Data-driven: splits credit across touchpoints based on past data, letting the system learn which interactions helped most.

Last-click: gives full credit to the final ad touch before the sale, which hides the value of earlier, awareness work.

Example journey: a user sees a search ad, later sees a remarketing banner, then returns and buys. Last-click will credit the last ad. Data-driven will split credit across both.

That means upper-funnel keywords or display placements may look useless under last-click, but show real value under data-driven.

• Rule of thumb: use data-driven unless you need strict, historical comparison.
• Warning: models don’t fix bad tagging — if your tracking is broken, the model only divides wrong numbers faster.

“If your tags are clean, data-driven attribution gives a fuller picture; if not, fix the tags first.”

Pick a single scoreboard for results so your bids chase real value, not noise. In practice that means only the actions you truly value should appear in the main “Conversions” metric. Everything else can live in “All conversions” for insight without steering bids.

“Conversions” is the metric used by automated bidding — it’s the one that changes spend. “All conversions” includes those plus softer events you want to observe but not optimise for.

Primary actions are those that link tightly to revenue or a clear sale outcome. Examples:

• Primary: purchase, qualified lead form submit, phone calls over a meaningful duration.
• Secondary: brochure download, pricing page view, time-on-site, add-to-cart when you’re lead gen.

Counting lots of micro actions as primary inflates your reported conversions. That makes cost-per-lead look great while real sales stay flat. Multiple micro events from the same user can paint a false picture of volume and quality.

Quick checklist: mark an action primary only if it ties to revenue, is unique per customer (not repeatable spam), and has a clear success signal.

Don’t trust numbers until you’ve proven the tag behaves the way you expect. Run quick, practical checks so reports reflect real customers, not accidental clicks.

Open google tag manager Preview, perform the conversion step on your site, and watch the debug pane. Confirm the conversion tag fired on the exact event you intended.

Use Tag Assistant as a second opinion to see what loaded on the page and catch duplicate or missing tags.

Use the platform’s “Test installation” option to sanity-check setup. Remember: test status can take time to update — don’t refresh every minute.

Expectation: actual conversions typically appear in reports after at least 24 hours. If nothing shows after that, re-run the Preview flow and check dataLayer events.

Look in campaign, ad group and keyword views. Focus on the Conversions column for optimisation and All conversions for softer signals.

• Open GTM Preview → perform the conversion → confirm tag fired.
• Run Tag Assistant to verify page loads and rule out doubles.
• Use Test installation then wait 24+ hours for report data.
• Check campaign, ad group, keyword reports and the Conversions vs All conversions columns.

“Testing is quick, and saves a lot of guesswork — do it before you trust numbers.”

Enhanced conversions are a privacy-safe way to improve match quality when cookies fail. You send hashed first-party details — like an email or phone — so the system can better link a sale or lead back to the original click.

The hashing step means data is encrypted before it leaves your site. Only send first-party details you legitimately collected and that users consented to. If you’re unsure, get dev help — this usually needs a dataLayer variable or a form hook.

Example: a lead form with an email or a booking form with a phone number. When someone signs in on another device, hashed data boosts match rates so the sale appears in reports more often.

Practical win: fewer lost attributions on long journeys and better learning for bidding tools without exposing raw personal details.

Consent mode lets you adjust which cookies and signals run until users opt in. It protects privacy and keeps partial measurement working so reports don’t drop to zero when consent is withheld.

Tip: configure consent controls so you preserve useful signals but respect user choice — don’t turn everything off by habit.

Turn off “Automatically detect user-provided data” in your tag setup unless you’ve implemented it intentionally. Auto-detect can send fields you didn’t mean to share.

If you’re not confident: ask a developer to wire hashed values into the dataLayer and test in Preview mode. It’s a small dev job that saves messy data and privacy headaches later.

You’ve got the map — now make it work.

Quick recap: define what matters, choose the right types, prep access, create the action, install the site tag, then test and enhance. Do those steps in order and your reports will tell a truer story of return on ad spend.

Once live, treat this as a simple routine: check, test, refine values, and keep primary conversions clean so bids chase real customers, not noise.

Want a second set of eyes? Book a free audit at loudachris.com.au or visit /contact for help.

FAQ

How long until conversions show as active?
Reports usually show test installations within 24–48 hours. Tag previews are instant, but real report data needs time to process and match clicks to later actions.

Should I use GA4-imported events or native tags?
Native tags often give cleaner, faster reporting for paid campaigns. GA4 imports help unify analytics but can delay or distort campaign-level attribution. Use native tags for primary actions, import for extra insight.

What conversion window should I pick?
Pick a window that matches your sales cycle — short (7–14 days) for fast buys, longer (30–90 days) for services with longer decision times. Align it with your typical purchase journey.

Why do conversions differ from All conversions?
“Conversions” drives automated bidding and should be your primary scoreboard. “All conversions” includes softer events for insight. Keep only revenue-linked actions in the main metric.

Do I need enhanced conversions in 2026?
Yes if you collect first‑party data and want better match rates. Enhanced methods improve attribution with privacy-safe hashing, but implement them only with consent and dev help.