5 examples how AI can beat rule-based Facebook automation

— Which of these frying pans you will choose to cook pasta? — I don’t want a pan. I just want to eat, please show me the nearest Italian restaurant!

Rule-based Facebook automation systems assume that you as a marketer want something specific. They say: you might want to increase the budget by 10% every day OR you might want to turn off ad sets with CPI lower than 2$.

It’s like saying you want to buy a frying pan to make a pasta at home when in fact you just want to eat. And going to a restaurant might be a better idea.

The truth is that a marketer wants better campaign performance and wants marketing goals to be achieved with less effort.

Why marketers should stop using rules to automate campaigns?

You may think that creating rules is a simple way to manage campaign performance. In other words you ask your Facebook campaign to behave the way you put in the rule. Yes, it will. Unfortunately it does not mean it will result in better performance.

• Rules can’t take into account different metrics simultaneously and find the best balance between scaling and budget efficiency.
• Rules are based on our feelings mostly. Why to choose 10% budget increase, not 15%? We never have a strong argument about that.
• Rules need to be changed. What worked yesterday might not work tomorrow. And you will need to go and change the rule again.

#1 Example — AI can estimate conversion rates better

How do you feel when creating rules for a new campaign? You need to estimate lot’s of metrics based on the results of your first test launch. If your goal is to drive app installs it’s important to know your installs to clicks rate, CTR, the minimum number of clicks before the first install etc. You can use excel, your brain or go to the 3rd party dashboard. But the truth is that you almost never set good working rules from the first time. Because the reality will change. Tomorrow you will need to do the same work. Or you can let AI do this work for you and analyze your funnel and conversion rates in realtime.

#2 Example — AI can better know how to compare results

How many times you thought about choosing the right time period to compare ad sets performance?

With AI you don’t have to thinks about time ranges anymore. It will automatically decide how to compare your ads and choose the best ones to scale using multifactorial models (when such factors as CPI, the number of installs per day, the number of days of statistics etc. are taken into account).

#3 Example — AI can predict your future campaign performance

Yes, many Facebook automation products use predictive analytics. They also have rule-based interfaces. They notify you about how your campaign will perform tomorrow, and you need to think about how to adjust the rules having this information. With AI you don’t need to reset your rules — it will adjust next steps (duplicating, budget increase etc.) automatically. The power of predictive analytics starts when it can be applied to your next steps in a couple of simple actions — like telling your virtual AI assistant which of them to implement right now.

#4 Example — AI can tell you which metric now is more relevant to optimize

You can set rules to manage your CPI. What if you have not enough actions to understand how to improve your campaign? How to decide when is the right time to change the metric?

Usually you should go and calculate your funnel to set new rules to manage CPC or CPM. And you need much effort to change these rules each time. With AI you will be sure you optimize the right metric without manual calculation. Your virtual assistant will suggest the next action for the metric that will give you the maximum performance.

#5 Example — AI can suggest how fast you can scale your campaign

Let’s say you set a 20% budget increase rule when CPI < 2$. Why?

You did this based on our feelings mostly. AI can help you understand how fast you can scale to balance between high volumes and an optimal cost per action.

How marketers can benefit from AI in the next 3 years?

Many of us still choose most popular solutions to automate Facebook ads. There are a lot of companies who solve this problem. Let’s think one step further. Facebook has already implemented it’s own automated rules which are pretty good comparing to 3rd party solutions. Now marketers need smarter AI to replace some rules and effectively manage the other. We will see new solutions and companies who will manage routine tasks and be able to analyze tons of data in seconds. AI will empower marketers to more fully utilize their uniquely human skills.

That’s why we created Leadza.ai — a virtual assistant who can help better optimize campaigns by sending daily optimization tips through Facebook Messenger. Our mission is to simplify the decision making process for any marketer who faces the task of managing ads and to make this work enjoyable.