How To Optimize Facebook Ads In 4 Steps
Why did one ad you created as an afterthought get you so many new customers?
Why did this other ad you’re actually proud of perform so poorly?
Optimizing Facebook ads has one main goal – improving performance over time.
Essentially, we aim to get more for less – We want to maximize results while minimizing the cost required to reach them and we do this by understanding why certain things worked while others did not.
This is how it’s done – the 4-step process to effectively optimize Facebook ads:
Topics covered in this post:
- What’s your success metric
- Step 1:Scale up the best, pause the worst
- Step 2: What have you done?
- Step 3: Perform data analysis
- Step 4: Test
- How often should you optimize?
Answer this question before you optimize Facebook ads:
What’s your success metric(s)?
You can’t optimize Facebook ads before you choose the key performance indicator (KPI) with which to differentiate successful ads from those that failed.
This KPI is your goal, so, for example, if you’re advertising on Facebook in order to get more sign ups / paying customers (=conversions) then the KPIs would be:
- number of conversions
- cost per conversion
You want your ads to yield as many conversions as they can, given a budget and a predetermined cost per conversion.
(* How do you determine an acceptable cost per conversion? My post – calculating customer lifetime value can help you with that)
If you’re advertising on Facebook in order to increase brand awareness (=getting more people to know your brand), your KPI can be Facebook shares.
If it’s community engagement you’re after (=building relationships with your existing fans and customers), focus on comments and Facebook reactions.
Step 1: Scale up the best & pause the worst
Now that you have your KPIs in mind, you can move on to optimize Facebook ads effectively. Steps 1 is going to be relatively easy:
You should already know what constitutes ‘good’ and ‘bad’ ad performance based on the KPIs you chose.
Now create a binary rule for yourself that states – ‘ads that have a cost per action of over X will be paused’, ‘ads that have a cost per action of under Y will be scaled up’.
If conversions is your goal:
Pause all ads that failed to reach your cost per conversion.
Ads that brought in many conversions for a lower cost per conversion than you chose should be scaled up. This means increasing their daily budget and giving them an even higher bid (if you’re bidding manually).
*If you have ads that have performed exceptionally well, duplicate them and use them to reach new [yet similar] Facebook target audiences.
If brand awareness is your goal:
Pause all ads that failed to reach the cost per share you want.
Ads (and by ads I mean sponsored Facebook posts of course) that were shared many times and achieved a lower cost per conversion than you chose should be scaled up. This means increasing their daily budget and giving them an even higher bid (if you’re bidding manually).
*Just like with conversions – if you have ads that have performed exceptionally well, duplicate them and use them to reach new [yet similar] Facebook target audiences.
If engagement is your goal:
Pause all ads that failed to reach the cost per engagement (=comments / likes / video views / other) you want.
Successful ads should be scaled up. This means increasing their daily budget and giving them an even higher bid (if you’re bidding manually).
*And again – if you have ads that have performed exceptionally well, duplicate them and use them to reach new [yet similar] Facebook target audiences.
Step 2: What have you done?
The first stage was a fairly technical one, now begins the analysis portion that’ll enable you to optimize Facebook ads effectively and learn why certain things worked while others didn’t.
If you created a content framework before you started to run ads, then you already know what you’ve done, but if not, you need to breakdown your ads into their main characteristics.
For each ad, list all ad features that could have affected ad performance, for example –
- Who did you target
- Which Facebook ad type you used (photo post / link post / video / etc.)
- What value proposition you focused on and how many times was it repeated
- Which CTA did you use
- Describe the main features of the visual you used (color / black and white, whether you see a person or not, was there text on the image, if so what did it convey, etc.)
- Post length (short / long)
- Did you put a link in your post or not
- If you sent users to an external landing page – did the visual on the landing page match the one in the ad
- If you sent users to an external landing page – did the message on the landing page match the one in the ad
- Anything else you can think of…
Step 3: Perform data analysis
To optimize Facebook ads effectively you need to derive insight from the abundance of data you have accumulated and utilize it to create better performing ads in the future.
This is by far my favorite step in the optimization process because it actually takes some skill to get right.
In step 2 you made a detailed list of the characteristics of your best, worst and ‘meh’ ads.
You now have information about all the target audiences you tested and the messaging and visual featured you utilized in order to attract them and you also have the ad performance results so you also know which feature-combinations worked well and which didn’t.
Now it’s time to analyze the data – make notes of any possible connections you see between features and results.
For instance, if you see that all the ads that used the call to action “Start Now” performed well, note it.
If you see that visuals with real people tended to perform better than cartoon/ animations, note that.
If one target audience responded better to a certain value proposition while another target audience responded better to another value proposition, take note of it as well.
And so on.
Step 4: Test
Once you squeezed all the possible insight you could from your ad data into notes about possible connections it’s time for the last part of your mission to optimize Facebook ads – testing.
These notes you gathered are essentially hypotheses of different elements that might affect ad performance and in order to learn if you were right and these features do contribute to ad success/failure, you need to test them out.
Create small-scale tests of Facebook ads based on the feature-combination insight and see what works.
And.. That’s it!
FAQ: How often should I optimize Facebook ads?
Optimization can be a long process if done thoroughly but that’s not always necessary.
I suggest focusing on doing step 1 on a daily basis and completing steps 2 to 4 on a weekly basis (and if not possible – at least once a month).
[An advanced option is automation and I’ll write a separate post about optimization automation soon.]
Have something to add to this post? Have any question or comments? Contact me here
Don’t forget to check out the homepage for some more How-To guides and insight into online marketing dilemmas