top of page

3-step Framework To Fix Your Ad Campaigns

Whenever performance decreases, we need to find the reason ASAP to know where to fix things.

The thing is, many times we assume we found the problem...

...but we end up fixing the wrong thing.

And there's nothing worse than fixing the wrong thing. Right?

It will only cause more issues.

That's why we have a simple 2-question system we ask every media buyer in our agency to answer in these cases:

  1. "Where's the problem?"

  2. "How did you realize it?"

These are simple, yet powerful questions.

How does this help?

The first question will answer where they think the problem is, obviously.

But, believe it or not, this question is not enough.

Why? We'll get there in a minute.

By asking "how did you realize it?", we're asking them to tell us what made them realize the problem was where they said it was.

Makes sense?

For example: I've heard things like "there's an issue with the pixel".

And when I asked them how they realized it, they were trying to find the evidence, and it turned out it wasn't the issue.

Now, although answering the second question is key for this to work properly, answering the first one ("Where's the problem?") is no easy task.

And without a process, media buyers could spend hours...

...when they could actually save all that time by following this 3-step framework.

Let's get started!

First, let's set up the column sets we need on the Ads Manager:

High-level overview:

  • Impressions

  • Frequency

  • CPM

  • Outbound clicks

  • Cost per outbound clicks

  • Outbound CTR

  • Landing page views

  • Cost per landing page view

  • Add to carts

  • Cost per add to cart

  • Checkouts initiated

  • Cost per checkout initiated

  • Website purchases

  • Conversion Rate (website purchases/landing page views)

  • NC-CPA (New Customer Acquisition cost) - See this in TripleWhale or any other attribution tool

  • NC-ROAS (Return On Advertising Spend) - See this in TripleWhale or any other attribution tool

  • Amount spent


  • CPM

  • Outbound CTR

  • Outbound Clicks

  • Outbound CPC

  • Thumb-stop ratio (3-sec video-view / impressions)

  • Hold-rate (75% video view / impressions)

  • Drop-off rates for the following ranges

  • 0% to 25% video view-rate

  • 25% to 50% video view-rate

  • 50% to 75% video view-rate

  • 75% to 100% video view-rate

Now that we have our column sets, let's look at the 3-step framework.

To make it easier for you, here are some recommendations before getting started:

  • Identify the period in which the performance dropped.

  • Open 2 browser tabs with the Ads Manager: one for the period in which the performance dropped, and another one with the previous one (in theory, when things were working well, or better than the current period). You can also use the "compare" feature, but this works best for me.

  • Order the ad campaigns by "Amount spent" DESC (having in the first row the campaign that spent the most amount in that period) to start looking at the campaigns that had more impact on the performance.

Now, let's get into our framework.

This framework consists of an ordered list of questions to ask yourself:

Step 1:

Ask yourself: "Is the problem before or after people land on the website?"

Check whether the Conversion Rate (CR) dropped or not.

If it did, you can assume the performance after the click is fine.

And thus, there must be something wrong before people land on the website.

On the flip side, if the CR dropped, there's something wrong after people land on the website.

Makes sense?

If the answer is "before"

Step 2:

Ask yourself: "Why did the CPC increase?"

If the CR was more or less the same in both periods, it means each visit was more expensive (unless you decreased your budget).

Then ask yourself: "Is it the CPM or the CTR?"

If you didn't know, CPC is a calculated field: CPC = CPM/1000/CTR.

This means that the CPC increases when the CPM is higher and/or the CTR is lower.

Assess which metric got worse (CPM or CTR).

If it was the CPM, let's see if it was something related to high competition...

...or if your ads are being penalized for some reason (in this case, you'll typically see CPMs of $60, $80, or even higher).

Most of the time, you'll notice a decrease in the CTR, though.

Step 3:

Now, ask yourself: "Is it my audiences or my ads?"

CTR = Clicks / Impressions * 100

If fewer people are clicking on your ads, we need to assess whether this is happening at the audience or ad level.

How can you realize that?

You'd need to find the ad set that is not performing well...

...and then check if all of its ads are underperforming, or only a few of them.

And then check if those same ads are causing the same issues in other ad sets too.

That's how you realize whether it's one or the other.

Now, ask yourself: "If it's my ads, what can I do?"

Use the Creatives column set I presented above to analyze the affected ads one by one, and come up with conclusions that help you improve.

If the answer is "after"

Step 2:

This means your CR is lower compared to the previous period.

Ask yourself: "Where is the biggest drop-off?"

In the "old days", we used to do this analysis on the platform.

But now we do it on the backend (ie. Shopify) or on the attribution tool.

We should analyze the following conversion rates:

  • Landing page view > Add To Cart

  • Add To Cart > Checkout initiated

  • Checkout initiated > Purchase

If you're using Shopify or some of the well-known attribution tools, you can see these partial conversion rates report without any issues.

Step 3:

Ask yourself: "Why is there a drop-off here?"

Once you which of the above-mentioned conversion rates decreased, you can do a deep dive and analyze the reason.

Sometimes you won't have the answer, don't worry.

The important thing is that you'll know where the problem is.

And then you can ask your client what you don't know or what is outside of your SOW.


Real Life Examples

Let me give you 2 real examples to put this into practice.

These were issues we've had with 2 different clients.

  • Problem #1: "sales are going down compared to last week"

  • Problem #2: "sales are going down compared to last week"

You read correctly.

At the surface level, the symptoms were pretty much identical in both cases.

And the truth is that most times this will be your red flag in the short term.

That being said, these similar problems were caused by two VERY different factors.

Let's take a look.


This happened to an activewear brand that was getting many orders per day.

All of a sudden, the number of orders dropped considerably.

We ran this 3-step framework:

Step 1: we noticed the CR had dropped considerably. So we focused on the "after the click" side of the equation.

Step 2: we found there was a drop in the Landing page view > Add To Cart rate.

Meaning, there were fewer people adding products to the cart.

That got our attention, but couldn't figure it out.

Step 3:

We visited their website and couldn't find some products in stock.

Then, we checked with the client, and it turned out they were out of stock of their best-selling products and forgot to tell us.

See how each of these 3 steps helped us, with simple steps, to realize where the problem was?


This happened to a multinational outdoor brand.

All of a sudden, the number of orders dropped considerably.

Guess what we did? We ran our 3-step framework :)

Step 1: we noticed the CR had dropped considerably. So we focused on the "after the click" side of the equation.

Step 2: we noticed the drop was in the Checkout initiated > Purchase stage.

We honestly didn't expect this.

We typically see these problems upper in the funnel.

Step 3:

We went to the website looking for some hints and found the issue.

It turns out there was a bug in the code of their store!

That bug was causing everyone who landed on the checkout page to see it in German...

...and the prices in Euros!

While their main market was the United States.

To our surprise, they didn't know about it.

So luckily now there were able to fix it, and we understood why things were getting worse.


What could have happened without this analysis?

In both cases, endless optimizations in the ad account.

Trying to make things work.

Maybe even fixing the wrong problems.

Wasting a lot of client money in the process.


bottom of page