10 minutes

How to Export & Analyze Facebook Ads Data Like a Data Scientist in 2024

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.

Leverage the Power of Facebook Ads Data

Introduction to Data-driven Marketing

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?

Extracting Valuable Insights from Your Data

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:

  1. Navigate to your Facebook Ads Manager.
  2. Select the campaigns, ad sets, or ads you wish to analyze.
  3. Click on “Export” and choose the format that suits your analysis tool.

Data Cleaning:

  1. Ensure that the exported data is free from discrepancies and null values.
  2. Use tools or scripting languages like Python or R for efficient data cleansing.

Data Analysis:

  1. Utilize statistical measures to understand data distribution.
  2. Employ visualization tools to observe trends and patterns.

Implement Insights:

  1. Apply the insights derived from the analysis to optimize your ad campaigns.
  2. A/B test different strategies and compare their performance.

Utilizing Data Analysis Tools

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.

How to Analyze Facebook Ads Data

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.

how to analyze facebook ads data

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.

Export Your Facebook Ads Data

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.

  1. Customize Your Columns
  2. Add Breakdowns
  3. Select Your Dates
  4. Export Your Data

Customize Your Columns

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”.


Facebook Ads Interface

For instance, if you’re managing an eCommerce store, you may want to include metrics such as:

  • Purchase Conversion Value
  • ROAS
  • Adds to Cart
  • Checkouts Initiated

Add Breakdowns

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.


Facebook ads breakdown

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.

Select Your Date Range for Analysis

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”.

Facebook ads date range

Export Your Data

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”.

how to export facebook ads data 1

Then, choose “Export as .xls” - or any other options, if you prefer. I prefer to uncheck the “Include Summary Row” for cleaner data.

how to export facebook ads data 2

Secret Trick to Adding More Dimensions in Your Analysis (Optional)

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.

Facebook Ads Account Naming Conventions

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:

  • Stage: Top of the Funnel (ToF)
  • Campaign Objective: Conversion (Purchase Event)
  • Budget Type: CBO (Campaign Budget Optimization)
  • Offer: 30% Off
  • Date: 15.11.2021

There are many additional “labels” (or breakdowns) you can add to your naming conventions. Here are a few ideas:

  • Creative Type: Images, Videos, Carousels, etc.
  • Audience Type: Interests, Lookalikes, Remarketing
  • Video Length: 15s, 30s
  • Landing Page: variant A, variant B
  • Creative Type: UGC, Testimonial, Endorsements
  • Any other segment you like

But how exactly do we use this information to our advantage? Keep on reading; we’ll show you how.

Use PolymerSearch to Automatically Identify Valuable Patterns

Upload Your Data to PolymerSearch

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.

upload to polymer search

We’re almost at the fun part.

Enrich Your Data with Polymer’s “Array Separator” (Optional)

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.


customize app settings

Then, click on “Columns Settings” and select the column you want to customize. In this case, “Campaign Name”.

customize polymer app

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.

split columns

With this change, we have now created multiple different columns as below: 

  1. Prospecting (ToF) – [Stage]
  2. Conversions: Purchase – [Objective]
  3. CBO – [Budget Type]
  4. Offer – [30% Off]
  5. 15.11.2021 [Date Launched]

Now, onto the fun part!

Get Automated Insights with Polymer’s Powerful AI

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.

AI Auto Insights

Use Polymer’s Auto Explainer for Automated Insights

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.

conversion value

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.

conversion value by gender and age

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.

facebook ads targeting

But that’s just the start of it. How can we dig deeper into our data?

Add Multiple Dimensions/Breakdowns

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.

gender and age demographics

Clearly, younger men don’t seem to respond well to the offer in this ad. However…

gender age demographics

It seems the same doesn’t apply for men +55 years old. 

Can we learn something from this information?

Calculate Your Own Metrics

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.

calculate ROI facebook ads

Next, we added “Creative Type”, “Media Type”, and “Angles” to our analysis.

high performing ads

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.

underperforming ads

These are only a few small examples of the type of insights you can find with PolymerSearch.

Enhance Your Marketing Strategies with Advanced Analytics

The Role of Predictive Analytics in Marketing

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:

  • Use clustering algorithms to group similar customers.
  • Tailor your marketing strategies to each segment, enhancing personalized marketing.

Customer Lifetime Value Prediction:

  • Employ predictive models to estimate the total revenue a customer can bring to your business during the entire relationship.

Churn Prediction:

  • Identify patterns that signal a customer is about to leave, allowing preemptive actions.

Implementing Machine Learning Models for Ad Optimization

Leveraging machine learning models can significantly optimize ad performance by automating the analysis process and deriving actionable insights. Here’s how:

Ad Spend Optimization:

  • Utilize regression models to determine the most cost-effective ad spend for each campaign.

Ad Content Optimization:

  • Implement natural language processing (NLP) to analyze which ad content resonates most with your audience.

Lookalike Audience Creation:

  • Employ classification models to identify and target new audiences similar to your existing high-value customers.

Note: Ensure to constantly validate and update your models for maintaining their predictive accuracy and reliability.

Conclusion

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.

Posted on
January 29, 2024
under Blog
January 29, 2024
Written by
Alex Almeida
Founder of Connected Brands, a performance marketing agency, Author at The Conversion Lift

Explore Our E-commerce Business Intelligence Templates

Browse All Templates

Related Articles

Browse All Templates

Start using Polymer right now. Free for 14 days.

See for yourself how fast and easy it is to uncover profitable insights hidden in your data. Get started today, free for 14 days.

Try Polymer For Free