Template
E-Commerce Business Overview
Combine your Shopify, GA4, Google Ads, and Facebook data for insights across your business.
.svg)
From the hundreds of available interests, audiences, campaign objectives, and dozens of available breakdowns, understanding how to properly set up and navigate the platform can feel like going down a rabbit hole.
The hardest part about analyzing Facebook ads data is finding the right target audience and figuring out which assets convert the best. There are hundreds of possible combinations.
In this article we'll show you how to quickly find meaningful insights in under 5 minutes using an AI-powered tool called Polymer Search.
In the era of digital marketing, making data-driven decisions is pivotal for success. Facebook, being one of the largest social media platforms, provides a wealth of data through its Ads Manager. But how can we effectively utilize this plethora of information to make informed marketing strategies?
Analyzing Facebook Ads data requires a systematic approach where the raw data is converted into valuable insights. Employing a methodological analysis can help in identifying patterns, understanding customer behavior, and enhancing ad performance.
Data Export:
Data Cleaning:
Data Analysis:
Implement Insights:
Data analysis and visualization tools like Tableau, Google Data Studio, or Power BI can be instrumental in comprehending your Facebook Ads data. Integrating these platforms with your data allows for a deeper and more visual interpretation of your marketing performance.
This article will be broken down into a few separate chapters that will guide you from start to finish. It shouldn’t take more than 5-10 minutes to read.
By the end of this article, you should be able to quickly export data from Facebook, upload it to PolymerSearch, and find some great insights you wouldn’t (easily) find in your spreadsheet.
Let’s get to it.
To start, you’ll first need to download your data from Facebook Ads into a spreadsheet.
You can do so by following a few easy steps.
The first step is to customize your data columns so you can include the most relevant metrics for your analysis.
To do so, access your Ads Manager interface and hit the “Columns” button in the main ribbon. Then, click on “Customize Columns”.
For instance, if you’re managing an eCommerce store, you may want to include metrics such as:
In the next step, you’ll want to add as many relevant breakdowns as you can.
In short, breakdowns are additional dimensions for analyzing your data, such as placements, dates, regions, and others.
To do so, you can click on the “Breakdown” drop-down and choose your preferred breakdown. Note that, in some cases, you can’t combine multiple breakdowns simultaneously, so choose wisely.
This is an essential step since this will give your data an added level of granularity which will help you find patterns you might’ve easily missed otherwise.
Next, you’ll want to select the date range where you’ll make your analysis.
Typically, you’ll want to choose a time window that’s big enough so that your analysis can be statistically significant. In other words, make sure you’re not trying to find patterns in your data from 10 clicks or $50 in ad spend.
To do so, in the top right of your ads manager, select your date ranges and hit “Update”.
The final step is the easiest: to export your data.
In the main menu, right next to the “Breakdown” button, click on “Reports” and then “Export Table Data”.
Then, choose “Export as .xls” - or any other options, if you prefer. I prefer to uncheck the “Include Summary Row” for cleaner data.
While this is an optional step, it's highly recommended because of the amount of data you can get from it you wouldn’t be able to otherwise.
As we’ve discussed, Facebook doesn’t give us too many options when it comes to the breakdowns you’re able to export. Plus, you can’t combine too many breakdowns at the same time.
So how do we go around this issue?
This is where account naming conventions come into play.
Naming conventions are important because not only do they allow us to navigate our accounts better, but they also give us extra granularity in the data we’re analyzing.
By adding a few “labels” to our campaign names, we can add more breakdowns than those given to us by Facebook.
Here’s an example of how we would name one of our campaigns.
Prospecting (ToF) - Conversion: Purchase - CBO - 30% Off - 15.11.2021
In the scenario above, we can easily tell multiple things about the campaign:
There are many additional “labels” (or breakdowns) you can add to your naming conventions. Here are a few ideas:
But how exactly do we use this information to our advantage? Keep on reading; we’ll show you how.
Now that you have your spreadsheet ready, it’s time to upload it to PolymerSearch so it can help you quickly identify patterns and trends in your data.
Head over to https://polymersearch.com and sign up for free. It takes less than a minute.
Once logged in, click on the “Upload CSV or XLS” button and choose your spreadsheet.
We’re almost at the fun part.
If you skipped the “Optimize Your Fields for Extra Granularity” section, feel free to skip this one too.
To access the information from your naming conventions, you’ll need to clean up your data, so it’s easily accessible to Polymer’s AI.
In other words, you’ll need to split your “Campaign Name” (or whatever other column you want to break down) with Polymer’s “Array Separator” feature.
To do so, click on the “Customize App” in the main menu’s settings.
Then, click on “Columns Settings” and select the column you want to customize. In this case, “Campaign Name”.
Finally, scroll down to the “Split the raw value as an array using a custom separator” and type in the separator you want to use. In our case above, the “-“ (dash) sign.
With this change, we have now created multiple different columns as below:
Now, onto the fun part!
Now we get to the easy (and fun!) part: getting meaningful insights for your business.
To start, once you’ve uploaded and customized your spreadsheet data, all you need to do is launch your app.
Then, in the main menu, enable the “Auto Insights” toggle.
One of the many great features in PolymerSearch is the Auto-Explainer tool.
In short, this tool allows you to simply enter a metric of your choice and it will automatically show you which columns have the highest outlier for that particular metric.
Here’s an example.
In the image above, we added “Purchase Conversion Value” as our metric and quickly discovered a few insights.
In moments, we can see that Polymer suggested that the column “Ad Name”, in particular the ad “image-lifestyle-monthly…”, drove significantly more revenue than the dataset’s average (+1,440%), compared to other ads.
By clicking on that specific ad - you can simply click on the label - we can easily filter our results so we can dig deeper into that particular dataset.
In this case, Polymer’s automated insights highlighted that women aged 35-44 were performing pretty well. Interesting!
Now, within another click, simply by adding gender to our breakdowns, we can confirm that men, particularly, don’t perform too well with this particular ad.
But that’s just the start of it. How can we dig deeper into our data?
Now, while the above may have been an interesting find, the real benefit Polymer brings on top of traditional spreadsheets is how easy it is to make multidimensional analysis within a few simple clicks.
As you know, spreadsheets with a lot of columns and rows can make it hard to pinpoint which segment of data is actually driving results.
With Polymer, that’s easy.
By adding “Age” and “Gender” as additional breakdowns, for example, we can now draw a few additional insights.
Clearly, younger men don’t seem to respond well to the offer in this ad. However…
It seems the same doesn’t apply for men +55 years old.
Can we learn something from this information?
Another great feature in Polymer is the ability to calculate your own custom metrics.
Now, let’s see what other insights we can find in our data set.
Instead of using our “Purchase Conversion Value” metric, let’s calculate our return on ad spend (ROAS) by adding “Amount Spent” to our “Minimize” field.
Then, let’s sort our metrics by ROI.
Next, we added “Creative Type”, “Media Type”, and “Angles” to our analysis.
Very quickly, we can tell that video product reviews, particularly about the marketing angle “clean food” are working out pretty well.
On the other hand, video testimonial reviews have accounted for 20% of our ad spend with literally no sales. Now that’s a problem.
These are only a few small examples of the type of insights you can find with PolymerSearch.
Predictive analytics involves utilizing statistical algorithms and machine learning techniques to identify the likelihood of future outcomes based on historical data. In the context of Facebook Ads:
Customer Segmentation:
Customer Lifetime Value Prediction:
Churn Prediction:
Leveraging machine learning models can significantly optimize ad performance by automating the analysis process and deriving actionable insights. Here’s how:
Ad Spend Optimization:
Ad Content Optimization:
Lookalike Audience Creation:
Note: Ensure to constantly validate and update your models for maintaining their predictive accuracy and reliability.
There are plenty of other use cases with Polymer’s AI, but if we were to explain them all, this would become an extremely lengthy post so instead you can check out these other posts:
We have written a separate post on finding winning Facebook ad creatives before, so feel free to check it out for some additional ideas on how to use Polymer.
See for yourself how fast and easy it is to uncover profitable insights hidden in your data. Get started today, free for 7 days.
Try Polymer For Free
.svg)
.webp)
.webp)
.webp)
