How-to: Understand and Report on Facebook Performance
Published: January 27, 2017
Author: Neeraja Koneru
Facebook reporting can be incredibly intimidating for newcomers to the platform; the variety of views and metrics available mean that pulling useful, actionable information is no easy task.
In this blog post, we’ll cut through the clutter, showing you how to set up custom columns for different campaign objectives and how to understand performance by device, ad set, and audience.
Let’s dive in.
1) Commonly Used Columns and Setup
Facebook offers a nearly overwhelming number of column options when it comes to reporting, and it’s often difficult to know the small differences between similar metrics. First, know that you can set up custom columns to view personalized metrics:
There’s a search option to quickly find the columns you need, and you can save the columns as a pre-set to easily reference later:
The kinds of columns you’ll want to look at will differ based on your campaign objectives.
Direct Response Objectives (Lead/Sign-Up/Revenue Goals)
You’ll want to look at user interaction that resulted in clicks through to the website; you’ll also want to track “events” on your site (downloaded whitepaper, filled out form, added product to cart, purchased, etc.).
A good set of starting columns is: Impressions, Link Clicks (unique link clicks if you’re interested in unique users and not users returning multiple times through your ads), Link CTR, CPC, Amount Spent, and Conversion Goal (purchase, submitted form, signed-up, downloaded something etc.).
Engagement / Branding Posts
Here, you’ll want to look at things like reach and impressions as well as reactions, comments, and shares of your post to see how users are engaging with the content on Facebook. Luckily, Facebook has a pre-made “engagement” filter you can use that will show you most of these metrics.
If you tag the posts with UTM parameters, you can also check through Google Analytics how these users have interacted on their website – metrics such as time on site, pages per session, bounce rates, etc.
This link can help you tag your posts appropriately so data will fill into Google Analytics.
2) Device Performance
Desktop and mobile ads understandably perform differently, so you’ll want to split desktop placements and mobile placements into separate ad sets. Bidding strategies are set at the ad set level, so segmenting by device is important to ensure you’re using the most appropriate bids and bidding strategy based on performance. Splitting devices into separate ad sets allows you to also better allocate budgets and easily report on performance by searching for and filtering by “Desktop” or “Mobile” in your ad set name.
3) Ad-Level Performance
Ad creative and messaging is essential to campaign success on Facebook, so it’s important to understand performance by ad creative. Low-engagement ads or ad fatigue could doom a campaign even if the audience being targeted has strong potential, so you want to make sure you’re positioned to identify and cut out these ads.
First, start by naming your ads appropriately so you can filter and see performance by image/messaging type.
When you “Edit” your ad, you’ll see an option to name it.
From there, you can filter by ad name to see performance:
Check ad performance frequently to ensure ad fatigue hasn’t set in and to ensure you’re showing the most engaging ads to your audience. Try testing different image themes to further understand what resonates best with your demographic.
4) Audience Performance
Looking at performance by audiences is one of the most critical views Facebook offers to understand how various targeting is performing. Audiences could be anything from lookalikes, to users who are “interested” in your competitors or relevant products/services, to email lists you want to target on Facebook. Basically, any targeting option you choose.
My preference is to segment out each audience into different campaigns to make reporting easier so again you can easily filter by audience name. For example, if you have three different types of lookalikes – email subscriber lookalike, high LTV purchaser lookalike, and abandoned cart lookalike, for example – I’d recommend splitting each type of lookalike into its own campaign instead of having one large lookalike campaign with many ad sets. The number of ad sets within a “generic” lookalike campaign can get unwieldy, and having separate campaigns helps maintain organization within the account.
You can view a campaign-level report in Power Editor to quickly see performance by audience type:
Overall, there are a lot of moving parts on Facebook regarding targeting options, and there are tons of reporting metrics you can view that can make it difficult to know where to look for meaningful data. Taking a look at ad creative performance, device performance, and audience performance is a good way to start to understand what types of ads and targets are performing well so you can appropriately optimize and scale.