How to Know Which Ad Is Working (And Which One Is Draining Your Budget)

You are spending real money on ads across Meta and Google, and both dashboards are showing activity, but you still cannot tell which platform is actually driving your sales.

This guide tells you exactly how to know which ad is working, what to measure, and how to stop making budget decisions based on the wrong digits.

Why Ad Performance Tracking Is So Hard to Get Right

Every ad platform lives inside its own walled garden. Meta Ads Manager shows Meta data, and Google Ads shows Google data.

Run campaigns on both simultaneously, and you will quickly find that both platforms claim credit for the same sale. A customer sees your Meta ad on Monday, does nothing. Sees your Google ad on Thursday, clicks and buys. Meta claims the conversion because the customer was within its 7-day attribution window. Google claims it too because of the click. Your actual revenue went up by one order. The platforms are reporting two.

This is attribution overlap, and it is the most common reason small businesses misread their campaign performance. According to Meta's own advertising documentation, default attribution settings count view-through conversions that most small businesses are not aware they are being charged credit for.

The Four Metrics That Tell You Which Ad Is Working

Most dashboards surface 20 to 40 different metrics. If you are trying to figure out how to know which ad is working, these are the only four that matter.

What is not on that list: impressions, reach, likes, engagement. These are visibility metrics. An ad can reach 200,000 people and drive zero sales. A high impression count does not tell you how to know which ad is working. It tells you how many people saw it.

Benchmarks to measure against:

According to Google's performance benchmarks, the average CTR on Google Search Ads across industries is 6.42%. The average Meta CTR is 0.9%. If your numbers are significantly below these, the problem is the creative or the audience, before it is the budget.

How to Know Your Best Performing Ad: Comparing Meta and Google Fairly

The mistake most business owners make is looking at each platform's reported numbers in isolation. Platform dashboards are built to present their own data in the best possible light. That is not a conspiracy, but a product design.

To identify your best-performing ad across platforms, you need the same metrics, measured the same way, in one place.

Align your attribution windows first

An attribution window is the period after someone sees or clicks your ad during which the platform claims credit for a conversion. Meta defaults to 7-day click and 1-day view. Google uses different defaults. Comparing the two without aligning these windows means you are comparing completely different measurements. Set both to 7-day click as a standard starting point, as recommended by Meta's Business Help Center.

Use your own data as the source of truth

Both Meta and Google track conversions using their own pixels and attribution models. Both have incentives to count as many conversions as possible. Your most reliable conversion tracking data is your own: Google Analytics 4, your CRM, or your e-commerce platform. Cross-reference what the platforms report against what your own systems show. If Meta claims 40 conversions but your Shopify dashboard shows 22 total orders, that gap is telling you something critical about how much to trust those platform numbers.

Pause one platform for a week

The cleanest way to find your best-performing ad channel is to pause one platform entirely for a week and watch your actual sales. This is called incrementality testing, and it cuts through attribution noise in one move. If you pause Meta and sales hold flat, Meta was probably not contributing much. If sales drop, it was working.

Ad Attribution: Why Your Numbers Do Not Add Up

Ad attribution is the process of assigning credit for a conversion to the right touchpoint. It sounds simple. In practice it is the hardest problem in self-managed advertising.

There are five main attribution models, each giving different credit to different touchpoints:

Last click - gives 100% credit to the last ad the customer clicked before converting. Default in most platforms. Overvalues bottom-of-funnel ads and ignores everything that built awareness.

First click - gives 100% credit to the first ad that introduced the customer. Overvalues top-of-funnel and ignores what actually closed the sale.

Linear - splits credit equally across every touchpoint. Feels fair but is rarely accurate to real customer behaviour.

Time decay - gives more credit to touchpoints closer to the conversion. Reasonable for short sales cycles.

Data-driven - uses machine learning to assign credit based on actual conversion patterns. Most accurate but requires sufficient conversion volume to work properly. Google recommends data-driven attribution for accounts with at least 300 conversions in 30 days.

For most small businesses, the practical approach is last-click attribution with a 7-day click window on both platforms, cross-referenced against your own sales data. Simple, consistent, and honest.

Conversion Tracking: The Foundation of Knowing Which Ad Works

If your conversion tracking is broken, nothing else in this guide matters. You cannot know which ad is working if you cannot reliably measure what counts as a conversion.

Three things to check before trusting any platform data:

Is your pixel firing correctly? Use Meta's Pixel Helper Chrome extension and Google's Tag Assistant to verify your pixels are firing on the correct pages. A pixel that fires on every page instead of the confirmation page only counts pageviews as conversions.

Are you tracking the right event? Tracking "page views" or "add to cart" when your goal is sales gives you data that looks good but means nothing for revenue decisions. Set your conversion event to the action that directly maps to your business goal: purchase, form submission, or phone call.

Are you deduplicating conversions? If you have both Meta Pixel and Google Tag firing on the same confirmation page, and a customer came from a Meta ad, both platforms may fire and count the same conversion. Use the Conversions API on Meta and enhanced conversions on Google to deduplicate accurately, as outlined in Meta's Conversions API documentation.

If you are running campaigns and nobody seems to be clicking your ads despite healthy impression numbers, broken conversion tracking is sometimes the culprit: the pixel is firing on the wrong page and showing inflated conversion data that masks a real performance problem.

How to Know Which Ad Is Working: A/B Testing the Right Way

Most small businesses A/B test incorrectly. They change multiple variables at once: different headline, different image, different audience, different budget. When one performs better, they cannot identify what drove the result.

Effective A/B testing means changing one variable at a time.

1. Test the headline first. The headline is the highest-leverage variable in any ad. A different headline on the same image to the same audience can double your CTR. Run two identical ads with only the headline changed. After 500 impressions per variant, you have meaningful data.
2. Test creative format second. Static image vs video vs carousel, with the same headline and audience. Format preference varies by platform. On Meta, video typically outperforms static for awareness. On Google, responsive search ads outperform manual text ads for most industries, according to Google Ads Help documentation.
3. Test the audience third. Once you have a winning creative, test it against different audience segments. The winning creative on the wrong audience still underperforms. The right audience with a tested creative is how you find a campaign worth scaling.,,When to call a test. Wait for at least 500 impressions and 50 clicks per variant before concluding. Calling a test too early on small samples is how good ads get paused and bad ones get scaled.

The Weekly Process for Staying on Top of Ad Performance

Once you have unified performance data in one place, the decisions become straightforward. This weekly rhythm takes 20 minutes.

• Monday: Check cost per result across all platforms. If one platform's CPR is more than double another's with no creative or audience changes, shift budget immediately. Do not wait until end of month.
• Wednesday: Check CTR on each active ad. If any ad has a CTR more than 30% below your best performer, pause it. You are paying for impressions that are not becoming clicks.
• Friday: Review conversion rate. Healthy CTR but low conversion rate means the ad is doing its job but the landing page is not. Before touching the ad, check load speed, headline alignment with ad copy, and call to action clarity. If your ads are not converting despite strong click-through, the problem is almost always after the click.
• Monthly: Review patterns across best performers. Look at your three lowest-CPR ads from the month. What do they share? Same headline structure, same creative format, same audience type? Your next campaign brief should replicate those patterns.

If you are managing ads without an agency, this weekly loop is the difference between campaigns that improve over time and ones that plateau and drain budget quietly.

FAQs

How do I know which ad is working if both Meta and Google are showing conversions?

Cross-reference both platforms' reported conversions against your actual sales data from your own syste: the sum of what the platforms report will almost always exceed actual sales due to attribution overlap. Your own data is the source of truth.

How long should I run an ad before deciding if it is working?

Wait for at least 500 impressions and 50 clicks per variant before concluding. For conversion-optimised campaigns, wait for at least 50 conversion events. Calling a test too early on small data sets is one of the most expensive mistakes in self-managed advertising.

What is a good ROAS for a small business?

Google Ads delivers a median ROAS of 3.52x. Meta delivers a median ROAS of 2.21x. For direct-response campaigns, use 3x as your minimum acceptable threshold, and adjust it based on your margins and product costs.

What is the difference between CTR and conversion rate?

CTR measures whether people are clicking your ad. Conversion rate measures whether people who clicked are completing the action you want. Low CTR is a creative or audience problem. A low conversion rate with high CTR is a landing page problem. Fix them in order.

How do I set up proper conversion tracking for my ads?

Install your Meta Pixel and Google Tag on your site, verify they are firing correctly using Meta's Pixel Helper and Google's Tag Assistant, and set your conversion event to the action that maps directly to your business goal. For more accuracy, implement server-side tracking via Meta Conversions API and Google enhanced conversions.

How do I allocate my budget based on which ad is working?

Sort all active ads by cost per result, lowest to highest. The ad with the lowest CPR is your best performer. Shift the majority of your budget there. Before making any major reallocation, check your small business ad budget split across platforms and make sure you are not starving your best-performing channel.

Knowing how to know which ad is working is not a technical skill but a measurement discipline. Track the right four metrics, align your attribution windows, cross-reference against your own data, and run a weekly review. That is the entire system.