Facebook's New Delivery Insights Dashboard (Part 1)

How To Use Facebook’s New Delivery Insights Dashboard To Identify & Rectify Audience Overlap

Written by: Joey Alter

One of the most common roadblocks to success for Facebook ad newbies or even for more experienced campaign managers is this:

Facebooks-New-Delivery-Insights-Dashboard-Part-1.jpgHow do I diagnose exactly what’s wrong with my campaigns?

And as a follow up:

Once I know what’s wrong, how do I fix it?

Well, I have some good news for every advertiser who wants to be more proactive about their Facebook traffic.

On April 6th, Facebook announced two big changes to help advertisers gain more transparency into how and why their ads perform the way they do.

The first is the estimated daily results graph, which looks like this:

estimateddailyresults.jpg

Those of you that have run Google AdWords traffic before may recognize this.

bidsimulator.png

Like Google’s Bid Simulator, the estimated daily results graph shows you a projection of how many more conversions you’ll receive per day when you increase your daily budget. The top of the blue shade is the high end of possibility, the bottom is the low end.

The second big change Facebook has rolled out is an updated version of the Delivery Insights dashboard you may have seen previously:

17119856_334424180293667_2811747821189857280_n.png

This new Delivery Insights dashboard provides a lot of really useful information when you’re trying to pinpoint what has gone wrong (or right) with your campaigns in the recent past.

To access the Delivery Insights dashboard, you’ll need to go to the adset level in your Facebook Ads account, then mouse over to the delivery column. You’ll find an option for “see delivery insights”.

Once you have the dashboard open, and click over to the “Auction Overlap” tab, you’ll get a series of percentages like this:

auctionoverlap.png

Auction Overlap is defined as the percentage of times the adset you’re viewing overlapped with another one of your adsets in the ad auction, and was automatically removed from the auction by Facebook.

Why is this bad you ask?

Adsets with high overlap rate lose chances to be delivered in favor of adsets with better performance. This in turns affects adset performance, and overall results. If you want to get a true test of your new target audiences, it’s important to take note of the overlap with existing target audiences in your account.

If you do find your adsets have a high auction overlap rate, you can solve this in one of two ways:

Consolidation

Simply put, instead of launching 10 adsets, each to their own interest, which may overlap, you can combine them into their own adset.

Instead of launching a lookalike from a custom audience of leads in one adset and a lookalike of an uploaded list of leads in another, you can include both in a single adset and avoid targeting overlap.

Targeting Refinement

This is the method I usually take as opposed to consolidation. There are some businesses that are more mass appeal, that can run really broad audiences and still get conversions and a return within goal.

But your offers or products may be a bit more niche, and have a smaller audience. In this case, it’s important to properly separate your targeting so you can get a real idea of what’s working (whether it’s this lookalike or that interest or this country etc.) and iterate from there.

If I’m targeting multiple lookalike audiences in one account, I start with the most valuable audience. Typically, I use a lookalike made from high value buyers (people who’ve spent a large amount of money).

I include this lookalike only in Adset A.

Then in adset B I take the next highest priority lookalike, let’s say in this case a lookalike made from leads.

Include the lead lookalike, and exclude the high value buyer lookalike to avoid overlap.

I do this for every lookalike audience I want to test, making sure to exclude each previous lookalike from each new test.

And then, if there are interests or behaviors that I want to test, I exclude all of the lookalikes I’m testing from the interest group.

This may seem like a tall order, but it’s necessary to do if you want your new audiences to get a real, true test, and not just be slotted out of the auction in favor of your historically performing audiences.

If you’re having trouble keeping track of your proper exclusions, I would recommend keeping a simple spreadsheet with the adset name in one column, included audiences/interests in another column, and excluded audiences/interests in a third column.

Avoid Auction Overlap Before It Starts

Proper campaign management starts with proper standard operating procedures (SOPs). Let’s make one today so you can avoid auction overlap in the future.

Before launching new campaigns:

  • Go to Audiences
  • Select all the audiences you’re currently running
  • Select all the audiences you want to run
  • Note Overlap
  • Edit the new saved audiences you want to test to either consolidate targeting or refine targeting
  • Go back to audiences and note audience overlap
  • Once adsets are launched check delivery insights to see the actual auction overlap rate

Get Enough Delivery So You Actually Know What’s Working

At this point, you may be thinking “Checking the overlap between every audience I upload is too time consuming” or “keeping track of all my exclusions is too confusing”.

It’s very important you take the time to minimize the overlap between your audiences, so they get enough delivery to get a true test.

This is where FunnelDash comes in handy.

You run adset A for a week, and it has good results. You check FunnelDash and verify the return on ad spend for that audience.

Then you launch adset B. Unfortunately, it has 50% overlap with adset A, and gets removed from the auction 50% of the time.

Because adset B has lost 50% of its potential impressions, that audience’s ROAS could be skewed down because it’s lost 50% of its chances to reach those key people who would opt in, and potentially purchase from your marketing funnel.

At this point, refine your targeting using the method above. Once your targeting is refined, then you can check FunnelDash again to find and compare the ROAS between audiences, and be confident that your audiences have gotten a true test.

Use FunnelDash to Monitor Your Next Facebook Ads Campaign

FunnelDash makes it easier for you to track where your leads and customers are coming from, and when they’re converting.

Right now, you can try FunnelDash, free for 14 days. Click on the button below to sign up for your free trial.

Stay tuned for Part 2 of this series. On Monday, I'll talk about what audience overlap means for your Facebook ads campaigns.

Leave a comment:

>