Online Entrepreneurs #03: Turning Data Into Dollars With Facebook Ads Analysis: A Real World Guide

Why is that most business owners and media buyers skimp on Facebook data analysis?

Beats me, because optimal media buying is all about isolating winning metrics and shifting spend.

I would recommend doing a thorough data analysis at least once a month.

No matter how well you think you know your market and your offer, this exercise is bound to throw up some surprises.

In today’s post, I am NOT going to share some bookish theories or fluff.

Instead, I will get down right to the brass tacks by sharing some snapshots/templates that I have personally used to analyze Facebook ad metrics for one of my ecom clients.

All the figures have been changed to protect client privacy, of course. So don’t read too much into them.

Rather, try to grasp the underlying logic.

Anyway, without further ado, let’s jump right into it.

One of the first things that you must  do, of course, is to identify your top performing audience by ROAS.

Like this:-

Using easy to understand naming nomenclature (like above) will really speed up your data crunching.

In the above example, the best performing audience is a 1% Lookalike Audience of the CRM database.

So, conclusively, I would explore viable options to grow my CRM database.

The other option would be to test out broader audiences like 2% LAL or 3% LAL, and its impact on ROAS.

Here are two more templates that I use to evaluate my best performing audiences.

Analyzing data by “ROAS” as well as “Cost per Checkout” is especially important.

Because, if you are experiencing a huge drop-off on the checkout page, you might need to work on your checkout copy or design layout.

And, while evaluating the top performing audiences is critical, analyzing your “Least Performing Audiences” can also throw up some interesting insights.

For example, if the CTR is abysmal, you can determine the type of creatives that you should avoid in future.

Next, let’s talk a bit about placements.

Off late, the Facebook algorithm has become really good at diverting spend to the best performing placements.

So, in all likelihood, automatic placements might work better than manual placements.

But, I would avoid jumping to conclusions, and try to make data-driven decisions as far as possible.

For example, I was recently analyzing ad metrics for a funnel selling info-products

The funnel is structured something like this: Main offer > Upsell 1 > Downsell 1 > Upsell 2 > Downsell 2

When I ran ads to just “Desktop”, the front-end ad costs went through the roof.

But the higher conversion percentage for upsell 1 more than made up for it.

See,

Many buyers like to make purchases (especially higher ticket ones)  on desktop because it “feels more secure”.

Illogical, I know.

But, as business-owners, it’s our job to unearth such fascinating aspects and then use them to make our marketing even more impactful.

And last but by no means least, I like to monitor the difference between “Unique Outbound Clicks” and “Landing Page Visits”.

If you are experiencing a huge drop off here, the underlying reason could be a slow page load speed.

Which, in turn, is pushing up your ad costs.

The best way to determine if you are lagging behind on this front is a competitor analysis.

How does your website page load speed compare to the competition?

I have been using GTMetrix for ages to identify optimization opportunities.

If you have a Wordpress site, and want to improve your page load speed, WP Rocket could be a good solution.

Hopefully, I have given you some food for thought.

If you have any questions or concerns about your Facebook ads, feel free to reach out.

Until next time!