So you have set up your
. You waited 14 days to allow Facebook to collect data and dynamically optimize. Looks like you’re ready to start manually optimizing your ads!
Below are 5 optimizations you can make right now to affect your learning and improve your performance, whatever your goal.
Once your ads have collected enough data for you to take action, you need to look at your performance broken down by demographic category. This will allow you to see performance against specific factors, including (but not limited to): age, gender, platform and placement.
As you can imagine, being able to see how much of your budget is being spent on which age groups and placements versus the performance of those age groups and placements versus the campaign/advertising average provides you with valuable opportunities to limit unnecessary spending in underperforming areas. Let’s say your main KPI is a target CPA of £15 or less, which most placements achieve, but the “Facebook Stories” placement consistently generates a CPA of £25 or more; in this case, manually excluding the “Facebook Stories” placement from this ad set will help drive the CPA down.
It is also interesting to use this data for other reasons than the optimizations you are doing here and now. If you notice a trend in all/most of your ad sets (for example, the 18-24 age group has significantly lower engagement rates), you can exclude them from future ad sets you create to ensure that budget is pushed to the highest performing areas from the start.
Budget changes are the core of your optimizations. Whether you are decreasing or increasing your budgets, there are a number of factors to consider. The first is your overall monthly ad spend budget: are you about to exceed your monthly budget? Often, this prevents me from increasing budgets, but when it seems that increasing spending on a certain set of ads will quickly generate more sales, I explain to my client the value of increasing budgets and get approval to exceed the agreed-upon spend. Second, you need to pay special attention to key performance indicators (KPIs). If your goal is to achieve a ROAS of 3 and an ad set consistently delivers a ROAS of 3+, you need to increase its daily budget to get more performance. Conversely, if a campaign/set of ads consistently delivers a lower ROAS than your goal, consider lowering its budget to limit unnecessary spending. At The Good Marketer, we stick to the 20% rule, which means that when we increase or decrease a budget, we only change it by 20%. Any overage will be considered a significant change and will return the ad set/campaign to the learning phase.
Analysis of ads
If you follow best practices, you should run 3-5 ads in each ad set. After 14 days of campaign operation, check the performance of each ad. If an ad is performing significantly worse than the other ads in the ad set, even though it has had significant spend (for me, significant spend is when an ad has spent between 1x and 2x of your target CPA), you may want to disable that ad. This optimization means more spending on ads that generate conversions.
This is a quick win optimization that guarantees better results, but while you’re looking at your worst performing ads, be sure to pay attention to your best performing ads. Are they videos, images, carousels, slideshows? What elements of your brand’s value proposition does the creative and copy highlight? This is all very useful information that should motivate future designs and advertising copy.
Duplicate your best performance
As mentioned above, you should increase budgets by 20% if an ad package is performing well. However, during key sales periods like Black Friday or Pay Day, you may not have time to slowly ramp up campaigns in this manner. This is where duplication comes in: if an ad set is performing well above your goal and you want to scale it quickly to maximize conversions, you can duplicate it. This is not an “evergreen” optimization (read: don’t leave it on all the time) but it is useful for short periods during high intent sales periods.
Breakdown by day
I put this optimization last in this list, but it is certainly not the least important. This optimization requires more time to get relevant data, as we will be looking at the best performing days of the week. Consumer trends can sometimes be unpredictable. If you can look at the data from the previous 5-10 weeks and see that Tuesdays and Fridays consistently perform better than the other days of the week, you should consider increasing budgets for those days and decreasing them for the other days of the week (to maintain your monthly budget).
This can be done automatically through Facebook’s “rules” or manually, by choosing to increase each budget individually or to increase the overall account budget.
I hope these optimizations have helped you discover immediate actions you can take to ensure your budget is spent in the right areas, to scale your campaigns and to improve performance.
So go grab a beer, sit down and start optimizing!