Why Your Ads Look Like They’re Failing: Setting Up Tracking That Tells the Truth

Professional header image for step-by-step guide: Why Your Ads Look Like They're Failing: Setting Up Tracki...

Your ads are working. Your tracking is lying to you. That’s a tough thing to hear when you’ve been staring at a dashboard wondering why your cost per lead looks so high, or why your inbox feels quieter than the numbers suggest. It’s the classic story: a campaign looks like it’s underperforming, you pull the plug, and then you find out later you were getting phone calls the whole time that never got attributed to your ads. You just couldn’t see them.

The truth is, the conversions you can’t see are the ones that make business owners kill campaigns too early. If you can’t trace a phone call or a form fill back to the campaign that generated it, you’re essentially guessing. And guessing gets expensive, fast.

This is a how-to for fixing that. We’ll walk through the simple setup that makes your tracking tell the truth, starting with why invisible conversions tank good campaigns and why your ad platform and your inbox keep disagreeing. Then we’ll get into the steps: setting up call tracking numbers, defining conversion actions that only count real leads, building a plain lead log you’ll actually stick with, and reconciling the platform numbers against your own records. By the end, you’ll have a clear picture of what’s actually happening and a much cheaper way to decide whether to stick with a campaign or cut it.

Why Invisible Conversions Make Good Campaigns Look Like Failures

You know the scene. The dashboard says you spent $600. The bank account agrees. But when you try to name one job that came from those ads, you draw a blank. So the campaign gets called a dud.

Here’s the problem. For a local business, the lead is usually a phone call or a form fill, not a website visit. If those calls and forms aren’t tied back to a campaign, the platform has nothing useful to report and you have nothing to trust. That leads to two expensive mistakes: cutting a campaign that was quietly producing calls, or doubling down on one that was producing tyre-kickers and wrong numbers.

The fix doesn’t need a big budget. It’s a simple setup most owners can run themselves: call tracking numbers, clean conversion actions and a plain lead log. We’ll walk through each step, and if you’re still planning your first campaign, our guide to Google Ads for local businesses covers realistic budgets and what to expect in 90 days. Just remember this is a measurement problem first and a spending problem second, so sort the tracking before you spend another dollar on clicks. A slow mobile site can also waste those clicks before anyone calls, so it’s worth checking why your small business website needs to load fast on mobile.

Why Your Ad Platform and Your Inbox Keep Disagreeing

If you’ve ever stared at a Google Ads dashboard showing 40 conversions while your inbox holds 12 enquiries, you’re not going mad. The two numbers are answering different questions.

Attribution windows are the first culprit. The platform can still credit a click for a set number of days after it happens, so a booking made today might be logged against an ad someone clicked weeks ago. A 30-day window is a common example, but check your platform’s own documentation for the exact default before you rely on it.

Then there are modelled and view-through conversions. Platforms may estimate conversions or count people who saw an ad without clicking. Those estimates never appear in your inbox, your phone log or your spreadsheet.

Definitions diverge too. The platform counts a submitted form; you count a booked job. Both get called a conversion, but they aren’t the same event.

And the numbers can drift in opposite directions. A form plugin might log the same lead twice, while a sale closed over the phone never gets fed back to the platform at all. If you’re weighing up platforms, our guide to Facebook Ads vs Google Ads for local Australian businesses covers which suits different business types.

The goal isn’t forcing the numbers to match. It’s knowing why they differ, and which set you actually make decisions on.

Step 1: Set Up Call Tracking Numbers So Calls Match a Campaign

Once you know why the numbers drift apart, the first practical fix is the phone. Call tracking means showing a tracking number to people who arrive from your ads, which forwards straight to your real phone, so you answer as usual while the campaign gets the credit.

Dynamic number insertion is the usual setup: ad visitors see the tracking number, everyone else sees your usual number, which keeps your Google Business Profile and printed signage consistent. That is how tracking vendors describe it, so treat it as a vendor perspective rather than a platform rule.

Here is the line to hold: a tracking tool counts every dial, but counting calls is not the same as knowing which calls mattered. That second job is yours, and we will get to it shortly.

To start cheaply, use one tracking number for your ads, keep your main number everywhere else, and make sure calls forward to the phone you actually answer, not a landline in an empty office. It is the same keep-it-simple thinking behind building a social media content calendar your local business will actually stick to. Before you change any settings, pull the current setup steps from your ad platform’s own help documentation, because those screens and defaults change and were not verified for this piece.

Step 2: Define Conversion Actions That Only Count Real Leads

Now that calls from your ads are being counted separately, the next job is deciding which actions deserve to be called a lead. Open your conversions list and audit every action currently firing, including the ones a past freelancer or template set up and nobody has checked since. Common culprits: contact button clicks, duplicate form submissions from the same person, very short calls, scroll depth, and newsletter signups you never follow up. Switch these off as primary conversions or remove them entirely.

The working rule is one conversion action per genuine business outcome. A completed enquiry form counts; a call past your minimum length counts. That way the account stops rewarding volume for its own sake. This matters because automated bidding follows your instructions. If it is optimising toward button clicks, it will happily buy more button clicks, and your cost per conversion will look fantastic while the calendar stays empty.

The number that matters isn’t cost per conversion. It’s whether those conversions are real business opportunities. For your enquiry form to produce real leads, it helps if your website gives people what they need before they submit it, which is the thinking behind what a small business website should actually include. Next, you need a place to record those real leads.

Step 3: Build a Plain Lead Log You Will Actually Keep

Now for the bit that keeps everyone honest: a plain lead log. It’s one simple record of every enquiry that comes in, so you have a version of the truth that no platform, tracking window or estimate can bend. Because each enquiry is logged once, no two campaigns can claim the same lead either.

Eight columns do the job: date, source or campaign, call or form, name and number, what they asked about, how it was handled, whether they became a customer, and a rough job value.

Keep the format deliberately boring. A spreadsheet is fine, a shared sheet is better, and having it open on the phone you answer is best of all. No new software required to start.

The habit that makes it work is filling it in when the lead arrives, not at the end of the month. Aim for about 30 seconds per lead, and store only the details you actually need.

Once the log has had time to build, it unlocks the number you can genuinely make decisions with: cost per real lead, being your ad spend for the period divided by genuine enquiries.

Get that running and you’re ready to line it up against the platform numbers.

Step 4: Reconcile the Platform Numbers With Your Lead Log

Now that your lead log is running, the last step is putting it next to the ad platform’s numbers, not instead of them. Compare like with like first: same date range, same time zone and same attribution rules on both sides. Attribution rules are simply how far back each system credits a click, and if one credits a click from three weeks ago while the other only counts today’s enquiry, you are comparing two different stories.

Match the lists on something both share: a lead ID, the phone number, or the date and time the call or form came in. Then line the two lists up side by side.

Expect a small gap. It is normal and not worth chasing. A large gap is the useful signal that something in the setup needs a look. Check the usual suspects: leads logged twice, jobs closed in person and never marked, after-hours calls that never reach the log, and spam enquiries inflating one number only.

Make this a monthly habit. Keep a short note of each discrepancy and what caused it. After a few months, the comparison takes about ten minutes instead of an afternoon. With the two lists talking, you can finally judge what a real lead looks like.

What Counts as a Real Call or Lead? Setting Your Own Bar

Once your log and platform numbers line up, the next question is the one no platform can answer for you: which of those calls and forms actually count?

There is no official rule for a genuine call. Google will report duration and whether a call connected, but it will not tell you what a “real” lead is. That is a judgement call, and the useful move is to make it deliberately and write it down.

A few sensible starting points to adapt:

  • Set a minimum call length. Some guides suggest 30 to 60 seconds; treat it as a starting guess, not gospel.
  • Decide how to treat repeat callers on the same number.
  • Note after-hours calls, which are easy to miss but may be genuine.
  • Exclude wrong numbers and spam. Every local business gets them.

Your bar is business-specific. A hypothetical: a 20-second call giving a plumber an address can be a real job; a 20-second call to a salon is usually a price check.

Then test it. Mark in your log which enquiries turned into paid work, and look back at how long those calls ran and how they arrived. That evidence beats any benchmark.

Finally, revisit your thresholds every few months. As your own data builds, refine the numbers rather than treating the first version as permanent.

Underperforming or Under-Measured? Read This Before You Cut a Campaign

You’ve set your bar for a real lead. Now use it to tell two very different problems apart.

Under-measured means your lead log is full of real calls and forms, but the platform shows almost nothing. That usually points to broken tracking, not a broken campaign. Don’t punish the ads for your setup.

Underperforming is the opposite: the log shows few real leads no matter what the platform claims. That’s a genuine campaign problem.

So don’t cut a campaign on platform numbers alone. My rule of thumb: check the log first, and give a new tracking setup 30 to 60 days of data before you judge it.

If the log confirms poor performance, work through these in order: targeting, the offer, the landing page, then phone answering. If the offer needs sharper messaging, our content framework for local Australian businesses can help you plan supporting social posts.

Sorting out tracking is the cheapest performance upgrade available before you spend another dollar on clicks. It costs more time than money.

Fix the Tracking Before You Spend Another Dollar

Before you touch the budget again, fix what you can see. Invisible conversions make working campaigns look dead. Platform numbers and real leads disagree for explainable reasons. A simple setup closes the gap.

This week: put a tracking number on your ads, strip out conversion actions that aren’t real leads, start a plain lead log today, and compare both sets of numbers at month’s end. If your ads need a copy tune-up once tracking is honest, see How to Write a Google Ad That Gets Clicked by Local Customers.

Before you cut anything, check the log, and give a new setup the fair run we talked about. No pressure, but if you want a hand, Epik Monkey works with local businesses on online advertising and websites, in plain English, for owners who do most of the work themselves.

Conclusion

Your ads probably aren’t failing. They’re just being measured badly. Invisible conversions make good campaigns look dead, and platform numbers will keep disagreeing with your inbox until you close the gap. A tracking number on your ads, conversion actions that only count real leads, and a plain lead log you actually keep will tell you the truth. Then let your first month-end comparison do the judging.