How to Audit a Marketing Funnel (and Find Where You Are Losing Money)

A funnel audit is one of the highest-leverage things I do, and it is also one of the most common test tasks you will get when you interview for a marketing job. The good news is that it is not complicated. You do not need fancy tools. You need your numbers, a calculator, and the patience to follow the drop from one step to the next until you find the leak.

In this post I will walk you through exactly how I audit a funnel. We will map the stages, calculate the conversion rate between each one, spot the biggest drop-off, figure out why it is happening, and decide what to fix first. I will use a simple example with real numbers so you can copy the process for your own product or a mock test task.

Map the funnel stages first

Before you can find a leak, you need to see the whole pipe. Write down every stage a person passes through, from the first time they see you to the point where they stick around and buy again. For most paid acquisition funnels it looks like this:

• Impression: someone sees your ad.
• Click: they click through to your site.
• Landing: the page actually loads and they view it.
• Signup or add to cart: they take the first real action.
• Purchase: they pay.
• Retention: they come back or renew.

Your funnel might have more or fewer steps. A SaaS trial funnel adds an activation step, and an ecommerce funnel might split checkout into cart, shipping, and payment. That is fine. The point is to write down the real path your customers take, then pull the raw number of people at each stage from your ad platform and your analytics. If you have not set that tracking up yet, my guide on UTM parameters and Google Analytics 4 will get you the counts you need.

Calculate the conversion rate between each step

Now put the numbers next to each stage and divide each step by the one above it. That step-to-step conversion rate is where the audit lives. A single overall number (ads to purchase) hides the problem, but the rate between two neighboring stages points right at it.Here is a simple example funnel from a $2,000 ad spend:

• Impressions: 200,000
• Clicks: 4,000 (2% click-through rate)
• Landing page views: 3,600 (90% of clicks actually loaded and stayed)
• Add to cart: 360 (10% of landing views)
• Purchases: 72 (20% of carts)
• Repeat purchase in 90 days: 15 (about 21% retention)

Write each rate out: 2% click-through, 90% landing, 10% add to cart, 20% cart to purchase, 21% retention. Now compare each to a rough benchmark for your channel and category. You are not looking for perfection, you are looking for the one number that is clearly worse than it should be. For a refresher on the metrics, see core marketing metrics and conversion rate optimization basics.

Find the biggest drop-off (your bottleneck)

The bottleneck is the stage where you lose the most people relative to what is normal. Do not just eyeball the biggest raw drop, because every funnel loses most of its volume at the top. Compare each rate to a benchmark and find the largest gap.In the example above, the 10% add-to-cart rate is the suspect. Getting 3,600 people onto the page and only moving 360 of them to add something means 90% look and leave. If a healthy rate for this kind of page is closer to 18-20%, then doubling that one step to 20% would roughly double carts to 720 and, holding the rest steady, take purchases from 72 to about 144. That is a doubling of revenue from fixing a single stage, which is why you always attack the worst rate first.One caution: make sure the drop is real and not a tracking artifact. A huge fall from click to landing view often just means slow load times or blocked analytics. Sanity check the numbers before you rebuild a page.

Diagnose the likely cause

Once you know which step is leaking, ask why. Each stage fails for its own reasons, so let the location of the drop guide your diagnosis:

• Low click-through: the targeting or the creative is off. Wrong audience, weak hook, or an offer that does not match what people searched for.
• Clicks that never become landing views: page speed. Slow pages and heavy images bleed people before the content even shows.
• Landing views that do not convert to signup or cart: message match and offer. The ad promised one thing and the page says another, the value proposition is unclear, or the offer is not compelling.
• Carts that do not become purchases: checkout friction. Surprise shipping costs, forced account creation, too many form fields, or limited payment options.
• Weak retention: the product did not deliver on the promise, or there is no onboarding and follow-up to bring people back.

In our example, the leak is at landing to cart, so I would look hard at message match and the offer. Does the page headline echo the ad? Is the main benefit obvious in the first screen? A read of what makes a landing page convert usually surfaces the culprit.

Prioritize fixes, then re-measure

You will find more than one thing to fix, so rank them by impact and effort. Impact is the revenue a fix could unlock, which you already estimated when you sized the bottleneck. Effort is how long it takes to ship. Rewriting a headline is an afternoon. Rebuilding checkout is a sprint. Start with high impact and low effort, because those wins fund the harder work.For our funnel, I would first tighten the landing page headline to match the ad and make the offer clearer, since that is high impact and low effort. Then I would run an A/B test so I can trust the result instead of guessing.Then re-measure. Give the change enough traffic to reach a real sample, pull the same conversion rate, and compare it to before. If the rate moved, lock it in and go audit the next worst step. If it did not, your diagnosis was wrong, so revisit the cause. This loop, map, measure, fix, re-measure, is the whole job. If you also run ads, pair this with my Google Ads account audit, which digs into the top of this same funnel.