To all my PPC expert colleagues, or to anyone who’s just curious about where Artificial Intelligence (AI) is leading us...this path we’re on right now is not as quite terrifying or exhilarating as you may think.
According to Frederick Vallayes, who wrote Digital Marketing in an AI World, everyone should reevaluate their understanding and refocus their attention on how to “Future-proof our roles as PPC experts.” We all know the world is constantly changing, and with machine learning in the picture, I understand the varying thoughts and opinions of what this means for us.
What may be frightening to a lot of us PPC experts is how much Google is able to automate our bids and campaigns.
Where does that leave us?
He lays it down from the very beginning of this book that machine learning certainly has its strengths, but when it comes to marketing, human intelligence overcomes. The foundation of intuition and creativity is what gives us the upper hand and helps prove that AI is not as destructive to our role as some may come to think.
We need to know what the Smart system is and isn’t doing. Vallaeys describes Google Ads machine learning as a black box, meaning we may not be able to tell certain factors that Google is taking into account when it comes to automated bidding. We need to focus on understanding the business level of technologies, not so much the technical, meaning our job is to determine if we need to direct the system to consider any missing factors that can make or break a business’s Google Ads performance.
Machine learning can’t do anything for a business unless we teach it what to consider. There’s a lot that machine learning doesn’t have, and one of those is intuition. AI is not at a point where it can consider everything we know and it’s not at a point where it can communicate a value proposition to another human being. That’s essentially what marketing is: creativity, communication, and human intuition.
As a PPC professional, we have to be the liaison between clients and machine learning. It’s our job to understand our clients’ goals and needs, which is where our instincts and marketing expertise come into play, as well as understand how to direct the machine learning’s focus and tweak it to achieve said goals.
Machine learning is great at doing its job of making calculations, collecting data, and finding patterns in data, but it’s our job to tell it how to obtain that data so that it can take a step in the right direction. We get the ball rolling by feeding the system the data it needs to bring success to businesses.
We help distinguish a business from its competition thanks to our creative side. We put together messaging and creativity together to set our clients apart, to highlight what makes their products or services desirable over others.
At some point, you may be wondering if the human element is completely replaceable. I’ve wondered as I read through the book and as I’m writing this blog … Vallaeys says he’s started to wonder a little, too.
Let’s just say, for now, that it’s too high of a cost to maybe replace the human factor and that it’s more effective for machine learning and humans to collaborate, rather than machine learning to replace us.
Nowadays, machine learning isn’t just about collecting data, but having the ability to make correlations from the given data.
Vallaeys gives an example on this with his experience of seeing a rabid raccoon running around Google’s campus one day. He wanted to file a ticket about this problem by going to Google’s internal website. Once he got into the ticketing system, it listed out different options for him to choose from that matched with what he needed. He had no idea where to file his ticket, so he made an educated guess and filed it under the security department. He quickly got a response from the correct team that helped him with this issue.
Let’s compare the process of machine learning and humans in this instance…
He says, “the ticketing system used a prediction engine to make this happen. ”The system looked through the history of all tickets that were submitted by employees and fed the machine learning model with the information. The system then looked at which department closed which ticket, and assumed that whichever department closed the ticket was the right one to deal with this rabid raccoon issue.
If Vallaeys had “called the kitchen staff, the kitchen staff might have reassigned it to facilities, then facilities may have reassigned it to another department, and so on and so forth.” Then eventually the case would have reached the right department.
Google’s model analyzed the ticket’s data and knew which department was the correct one. This example gives an idea of how powerful machine learning is. It looked through historical data to get Vallaeys the right department very quickly.
When it comes to scalability and complexity, Google takes the win. Creativity is still needed, and that’s where AI loses.
So then how should we position ourselves with machine learning in the picture?
Vallaeys discusses three different roles: The “doctor,” the “pilot,” and the “teacher,” which I will discuss in my next blog.
I’m looking forward to diving into Part 2 of this series with you all.
We'll get back to you within a day to schedule a quick strategy call. We can also communicate over email if that's easier for you.