Analyzing Facebook Ad Results and Optimization

The key to having success with Facebook ads lies with analyzing the results, making adjustments, and killing ads when they are not working. Secondly, you will want to ramp up the campaigns that are working.

Begin by identifying variations that you want to look at targeting. You can start with a general advertisement and go from there. In fact, you should include a general advertisement in your campaign in case the variations you imagine that might be required really aren’t valuable. Using the example of a slot machine game, slot machines are popular in multiple countries such as the United States, Canada, Australia, Hong Kong, Japan, and the United Kingdom. At first glance, you might think that different ads need to be set up for say Japan and the United States. And sometimes this is true, but it’s not always true. Sometimes you can hit on an ad creative that just works. The only way to find out is to test, test, and test.

Continuing with the slot example, we can make a single ad with our favorite video creative and then simply show it across the board. Then we can set up ad campaigns for each individual country and then run those alongside it. At the week’s end, we compare results. If there is no distinct advantage to specialized ads, we can kill them. If they perform worse, we definitely kill them.

The specialized ads might work in some cases but not in others. So maybe the same ad works in Japan and the United States, but it doesn’t work in Hong Kong. In that case, then you’ll want to work on tailoring an ad specifically for Hong Kong.

There may be other reasons to keep your campaigns separate, but for the most part, Facebook will let you break out reports that will let you break down things by location or gender. However, one reason that you may want to keep locations separate even though they may perform the same with the same interest targeting and ad creative is that costs are different in different countries. This also means that rewards are different in different countries. You can take this same lesson to the bank for different states and cities in the United States too.

Let’s have a look at it. The cost per acquisition may be about the same if you’re running the same exact ad across multiple locations. But the rewards might be different. For example, in mobile games, you earn money through ads that you show inside your game and through in-app coin purchases. Some locations might tend toward more coin purchases than others. Japan is better than Hong Kong, and the United States is where the money is (relatively speaking). However, advertisers pay different rates based on the country. They may pay a lot more in the United States than they do in say Hong Kong or France. That means that the cost of acquisition needs to be lower in those countries.

Two things you need to look at to see if your campaigns are working are:

  • Cost of acquisition of a new customer
  • The average lifetime value of a user

So, if I have a mobile game and the cost of acquisition of a customer is $0.90 and that customer will generate $2.50 in revenue, then that is a winning proposition. However, bundling locations in ad campaigns can cause problems. For example, in the United States, an average user might cost $0.95 to acquire, but they might on average bring in $4 in earnings. In Hong Kong, they might cost $0.85, but only bring in $1.25 in revenue.

The same thing might be in play for our real estate ad. You might easily acquire prospects from New York who are interested in moving to Arizona, maybe as easily as you’re getting from people in California. But California is a lot closer, so in reality, it’s a lot easier to make the move. So over time, you might find that you’re making a lot more revenue from California residents but only occasional sales from New York residents. The sales from New York still add to your bottom line, so you don’t necessarily want to cut them out. But maybe an approach you will take is keeping the campaigns separate and running a campaign aimed at NY residents that has a much smaller daily budget.

Testing Your Facebook Ad Campaigns

So, you’ll need to figure out what variations you’re going to run on your campaigns. It can be varying demos, interests, or just changing up creative. The goal is to throw everything out there and then find out what works and what doesn’t.

You can find out what works by running the campaigns for a week with very small budgets. In my opinion, below $3 per day is too small, but $10 a day is too large. The more you spend the more information you’re going to get, but about $5 a day for a week should get you enough information to determine if a particular video or ad copy is working for you.

Data to Check on Facebook Ads Manager

You can determine whether an ad is working for you just by looking at a few parameters. The main ones to check are:

  • Amount spent
  • Impressions
  • Clicks (or installs or page likes) 
  • Cost per result

You will want to know the click-through rate as well. This is simply clicks/impressions * 100. You can compare click-through rates between various ad campaigns and determine which ones are working and which ones are not working as far as generating the response to go to your site or page, to take that initial step of taking action.

Of course, cost per result is an important parameter. If you are in a new business, and you really don’t know the lifetime value of a customer, then you’re not really going to be sure if a given cost per result is acceptable or not, however, what you can do is compare across ad campaigns to see what’s working.

Now if you go to the Columns button and select Delivery, you can get the cost per 1,000 impressions of the ads. It’s important to compare this cost across campaigns as well. However, note that one impression doesn’t mean one person. The number of people is given by Reach. Some people might get more than one impression for whatever reason.

Another parameter to look at is Columns ??Engagement. This will let you see the cost per click. Targeting and Creative will bring many of these measures together in a single display, showing you the cost per result, cost per 1,000 impressions, click-through rate, and cost per click.

Even if you set up all your ads with the same exact budget – and that is what you should do – you might find some ads get a lot more impressions than others. This might happen because of demographic choices or interests that you’ve selected. If an ad is really sluggish with impressions, that is telling you that you will need to open it up a bit – possibly widening the demographics, location, or adding more interests (or reducing the interests).

Killing Your Facebook Ads

After the five-day period is up, it’s time to kill all the ads that are not working. Simply shut them 

off. This can be done by toggling the button on the far left of the listing. Keep in mind that you can turn things off at every level. So, if you have multiple ad sets and ads, be mindful of what you are really shutting off.

Future Steps

The succeeding step is the ramp-up phase. This can work in two ways, the first is mandatory. To ramp up, begin increasing your daily budget. Never increase by more than 15% per day. So, we will go like this – assuming that the trial period was set to $5 Daily Budget.

  • Day 1: $5,75
  • Day 2: $6.61
  • Day 3: $7.60
  • Day 4: $8.74
  • Day 5: $10.05

How high you go after this depends on how many ads your running and what kind of daily budget you can absorb. However, if you’re after a large number of prospects a $10 budget probably won’t cut it, and you’ll need to keep increasing until you get to $40 and above.

A word of warning about budget increases. For some reason, the Facebook system seems to respond better to gradual budget increases. So, you aren’t going to want to move from $5 to $40 or $100 in one shot. You will probably find it doesn’t work. Take it gradually – soon enough you’ll be at your spending goal.

Issues with Facebook Ad Bid Caps

Bid caps can work in a funny way. You might find – but this depends on a huge number of factors so maybe you won’t – that setting a bid cap you hardly get any traffic. In one test I did, using app installs where bid cap is replaced by an equivalent cost per install – I set a low price of 75 cents. Not much happened. I kept raising the daily budget.

It finally worked – but I had to set the daily budget at $400.

That might sound scary – it isn’t. At $400 the campaign was still proceeding at its glacial pace. It opened it up enough, so it brought in 20-30 installs per day, and I was happy to have them at that low cost per install.

This may happen to you or it may not, since there are so many factors that will influence how many impressions a given ad campaign is going to get. However, you can test the high budget if you are running into something like that which just isn’t doing anything. Just keep a very close eye on it so that you can shut off the campaign if it starts spending in an out of control fashion.

Cloning of Facebook Ad Campaigns

Years ago, cloning showed much promise when Dolly the sheep was born. We haven’t been overrun with clones, but the good news is you can clone your Facebook ad campaigns. They don’t call it cloning, they call it duplicate. It’s a tool that can come in handy to ramp things up.

Suppose you have a campaign with a daily budget of $20 and you want to spend $100 a day, and that campaign is working really well. Instead of gradually raising the budget, you can make five copies of it and run them simultaneously. That will help you get from there to hear a lot faster. It’s a technique that has worked for me many times.

Leave a Comment