Reactions, comments, shares, likes, etc. are the lifeblood of a social media post. You’ll want to monitor those comments, especially, as this is an opportunity to answer questions or address negativity.
The Facebook Ads Optimization Checklist
Assuming an individual ad set has 5 or more ads, look for the loser in the bunch, and pause it. From there, introduce a new ad. Repeat this monthly, or every $2,000 spent on the ad set.
Ad set targeting
There are several key ad set targets you should consider for any campaign:
Lists and Lookalikes
Target your email lists, Facebook engagers, Instagram engagers and website visitors. Also create 1% – 5% lookalikes for each. Target all of these in the same ad set UNLESS your list sizes are in excess of 100,000, in which case you can put them into two different ad sets.
Pick a series of relevant interests, job titles, employers, etc. that suit your target audience. Consider narrowing your audience using the ‘Must also match’ feature if it makes sense. But keep your audience size over 100,000 if at all possible.
This targets everyone within a certain geographic region. No other restrictions. This will create your largest audience, of course. It may seem foolish, but so long as you have a good objective for your campaign, it can easily become your top performer. It allows Facebook to show ads to absolutely everyone in your geographic area, find out who actually converts, and then build its own “behind the scenes” target audience. The advantage to this ad set is it automatically makes use of Facebook’s trillions of data points to create your ultimate audience.
Ad set duplication
If an existing ad set has a large audience size (e.g. > 1 million) consider duplicating it. Facebook will reconfigure itself and pick a different subset of people within that same large audience. Basically, it tries again, and might come up with better results.
Breakdown by Delivery
A caveat for using breakdowns – be sure that you’re looking at a large enough data set (i.e. date range) to make statistically significant decisions.
If you look at a breakdown by Placement, you can see at the ad set or ad level if there are certain placements that suck. This only really works if your goal is traffic, however, as you can see which placements are too expensive. If your goal is leads or conversions, Facebook won’t show you the breakdown by placement, so be very careful turning off an individual placement. You never know if it’s high costs are bringing in all the results.
As with Placement breakdowns, this will show you performance by country, region, or city. Same caveat applies, however, if you’re aiming for anything other than link clicks.
This one will sometimes show you conversions by age group, so it can be useful for removing certain age groups that aren’t performing.
This one also shows you conversions by platform (sometimes), so can be useful for removing poor performing platforms
Time of day
This will allow you to see if the majority of your results come from certain times of day. If so, you can create an ad schedule to only deliver your ads during peak times.
This is a fairly significant change. It involves creating a new campaign or duplicating an existing campaign, and using a different objective. You can switch a traffic campaign to focus on landing page views instead of link clicks, or take a conversion focused campaign that isn’t generating many conversions and change its objective to traffic, for example.
Consider the ‘Fast Takeoff Method’ for new campaigns to get them through the learning phase more quickly. This is where you use a higher daily budget than you plan to, until you get 10,000 impressions. After that point, drop your budget to where it would normally be.
Use organic content for new ads
Consider using an existing organic post as an ad. If it’s already getting good engagement, it could be an ideal candidate to become an ad. Just be sure the content makes sense for your campaign.
Beware the learning phase!
As a rule of thumb, don’t edit good performing ad sets or ads. If you want to try something new, create something new. If you edit something in Facebook, it will likely return to the ‘learning phase’. Facebook then needs to capture another 50 results to re-learn what works. You can completely mess up the performance of a great ad set by making a seemingly minor tweak.
That said, feel free to edit ad sets and ads that aren’t doing so well. Not much harm to be done there.