Futureproofing Your Role As A PPC Expert
The following is an excerpt from the book Digital Marketing in an AI World: Futureproofing Your PPC Agency by Frederick Vallaeys.
Is the sky really falling? It sure seems to be. Announcements about AI’s increasing importance in Google Ads have many PPC agencies and professionals confused, scared, and ducking for cover. After all, Google is by far the biggest player in pay-per-click (PPC) advertising. And yes: artificial intelligence, and specifically machine learning, are transforming the PPC landscape. But no: this does not mean that machines are ready to take over from us humans. In fact, the best results may happen when machines and humans work together.
There are many ways to futureproof your role as a PPC expert. However, the current changes do require you to reevaluate and refocus, and this book will tell you how.
When Google introduced AdWords in 2000, everything was done manually. Over time, Google layered more and more automation into the process. Working there from 2002 until 2012, first as a Product Specialist and then as the AdWords Evangelist — spreading the word about this new form of advertising — I saw all this happen from inside the company that invented online advertising as we know it today.
In 2016, Google CEO Sundar Pichai announced we would soon be living in an “AI-first world,” and that Google was going to become an “AI-first company.” Since then, machine intelligence has driven the changes that have produced a new generation of AdWords, which was rechristened Google Ads in July 2018. The push towards AI has many, perhaps most, PPC agencies and professionals wondering: what does this mean for me? Is the “AI-first” Google going to put me out of a job? Will the things I’ve been doing to make accounts successful, such as optimization and reporting, be handled by machines?
PPC angst is even stronger because so many digital marketers are fairly young, towards the beginning of their work lives. They may have thirty years of their careers left, and, in that case, what does the AI push mean for them? If you believe Google’s official pronouncements, machines will be doing everything long before it’s time to retire.
In this rapidly changing environment, the half-understood “artificial intelligence” jargon being thrown around only increases confusion. The media, seeing a trend, is now extensively covering this new, “disruptive” technology. Media dramatization of AI always seems to come down to the image of a beautiful humanoid robot — like Alicia Vikander in the movie Ex Machina — taking over from us poor humans.
However, the story of what’s really happening is considerably less sexy — and less frightening. Machine learning is great at figuring out correlations within masses of data: indeed, it’s a lot better at finding patterns in large data sets than we humans are. What this really means is we’re seeing the increasing automation of the highly repetitive, boring, and tedious tasks of pulling data from reports and looking for signals within the noise. This is where machines, rather than people, excel.
Marketing, however, remains a fundamentally human field: communicating with other people, telling them what you can offer, and inspiring them to act. Intuition and creativity are at the foundation of much of our human intelligence. We’ve done pretty well as a species for quite a long time, and machines are not going to suddenly become better than humans at communication or creative intuition.
Yes, machine learning is having a considerable impact on PPC marketing. This has been true for some time and will be even more so in the future. At the moment, however, it appears that we’re at an AI inflection point, which means it’s more important than ever to separate fact from paranoid fiction and figure out what’s really going on.
As one of Google’s first 500 employees and its first AdWords Evangelist, I have great respect for the company and its products and services, some of which I helped create. However, I also feel they could do more to explain how humans and machines can successfully collaborate. Without guidance from the company most responsible for driving the AI revolution, they allow paranoia surrounding AI and PPC advertising to persist. Yes, CEO Sundar Pichai has said Google will be an “AI-first” company. What tends to happen in big companies is whatever the CEO says becomes gospel: this is what we’re going to do!
This can reach the point of the absurd, as a Wall Street Journal article has described. At a breakfast meeting, a CEO commented that no blueberry muffins were available, although there were plenty of other things to eat, and he didn’t particularly care for blueberry muffins. He was just making small talk about the food. Blueberry muffins began appearing at every breakfast meeting thereafter. It was several months before the CEO realized that a very casual remark on his part had been taken as a hard-and-fast requirement.
What’s happening with Google Ads, now that the CEO has said they’re an AI-first company? Two years ago, during Google’s annual marketing event at which new products are launched, a product manager got up on stage and said: “In the past, you had to worry about which ads to write for which audiences. But now, our machine learning is smart enough that you can just submit fifty different ad variations, and the system will figure out which is the right one to show to each individual user.”
The speaker’s concrete example was a user looking for hotels. Google would already know if that particular user responds better to ads for inexpensive or for five-star hotels, or prefers results featuring a variety of different kinds of accommodations. The system then delivers the hotel ads the user will be most likely to click on. Since Google makes money from every click on an ad, there is a strong incentive to show the most clickable ads to every user every time a search happens. Hotels, of course, benefit too, because if the ads connect better with users, the rate of bookings per click — the conversion rate — is likely to be higher too.
In the last year, this process has intensified with the introduction of Google’s new “smart” features: smart bidding, smart campaigns, smart shopping. “Smart” is basically just Google’s new nomenclature for anything that’s machine learning driven. Now, instead of being told that the machine will figure out which ads to show, as was the case two years ago, PPC professionals are being told: “Tell us how much money you’re willing to spend, give us the site you want to drive the customer to, and let us handle everything else.” The smart system will figure out how to get you the most conversions for your money.
The increase in how much Google is able to automate frightens many PPC professionals. In the past, it was the agency’s job to figure out targeting, keywords, audiences, and the bidding process: how much the client should be willing to pay for each click. You would also have to determine negative keywords, bid adjustments, and much of what Google is now handling automatically in their “smart” campaigns. While these are important decisions, many are a function of math and statistics, which makes this element of PPC a prime point of automation. PPC pros, who are capable of much deeper thinking and strategy, should look in the mirror and ask what kind of PPC pro they really want to be.
From Google’s perspective, this shift is quite real. There is an absolute move towards increased automation. But what this actually means is unrelated to the messages about AI the media tends to promote. When the media talks about artificial intelligence, it’s generally not in the context of digital marketing. Instead, you’ll probably hear stories about IBM’s Watson doing a cancer diagnosis. Will AI replace human doctors? (Hint: very unlikely.) Or, you hear about artificial intelligence getting really good at playing certain games, having mastered chess and, more recently, the still more complex game of Go. People envision a human sitting across the table from a machine that has all the human’s capabilities and then some. This perception of AI is inconsistent with reality. People think that AI can — or will soon be able to — figure almost anything out much more easily and accurately than a human can. However, the reality today is AI can help humans get better insights more quickly, enabling smarter decisions.
As of early 2019, Google Ads “smart” machine learning solutions still tend to be very specific. They are point solutions, as is the case with AI in other industries and verticals. Smart bidding is all about bid optimization. Smart campaigns are about automatically figuring out ad targeting and how to allocate bids and budgets across different ad formats. Popular perception, bolstered by the media, seems to indicate that there is very little left for humans — you — to do. However, the company still offers several types of Google Ads campaigns in addition to “smart” ones. It’s for the PPC professional to figure out if a smart campaign is really the way to go in a specific situation, and the answer is sometimes no. Well-managed but more manual campaign types, built on long-standing optimization principles, frequently still work more effectively.
If you decide to go the smart campaign route, it can seem there will be little hands-on management to be done after launch. Even with smart campaigns, though, you can almost always drive better results by setting up multiple campaigns with different goals and targets. Google itself acknowledges this in its documentation, which includes provisos such as: “you should have different target returns on ad spend for different campaigns, possibly based on promotions you may be running.”
A hotel, for instance, may be running a reduced-rate special on what would otherwise be a slow weekend. While machine learning does what it does quite well, it may not realize you’re running a special promotion for two days this coming weekend quickly enough. By the time the “smart” campaign figures it out, your sale is over. You’ve lost the opportunity for increased sales because the smart system didn’t adjust your bids quickly enough. Human intervention was required.
The point is, even with “fully automated” solutions, there’s still quite a bit of management involved. When PPC professionals add their own human experience, intelligence, and creativity to the “smart” system’s recommendations, you get better results.
It’s true Google is doing some amazing things using machine learning to make its PPC customers more successful. At the same time, people are worried about what Google might “really” be doing, and especially about where innovations in machine learning are leading. A PPC professional’s first question is, “Am I going to lose my job?” An agency’s question is, “Are we going to lose all our clients?” The answer is: “You’ll have a job (and an agency) if you have the right skills.” However, you need to learn what the right skills are — ones that machines can’t replace — how to develop them, and how to position and market yourself and your agency. You’ll also have a job, because machines plus humans can outperform machines acting alone. The machines can achieve average results without you in the mix, if that is what you aspire to. If you want stellar results, a strong process combining the best human with the best machine skills is required.
Most people come to the right conclusions if you give them the right information. One of Google’s key tenets is transparency and information-sharing. The company’s attitude is that trends don’t need to be spelled out, they can be figured out. If you give intelligent people the same information, they’ll generally come to similar conclusions.
Information-sharing is the reason I decided to write this book. PPC professionals need to find out what AI really is and how it really works. Once you begin to better understand the technology and its impact on PPC advertising, you’ll be able to see your options far more clearly and can develop your own vision of how to benefit from, rather than be intimidated by, the technology. You’ll realize how you can skill or re-skill yourself not only to survive but to thrive. The PPC professional who takes the proper action will still have a job.
Then how, as an agency owner, can you also make sure your clients will continue to use your company’s services? You may be great at your job, but your clients are exposed to the same media you are, and many businesspeople are eager to try out whatever is “new and improved.” A long-time client might come to you and say, “Well, I just read about Google smart campaigns in Wired. Why aren’t we doing that? Why do I still pay a team of humans to do all these things, when artificial intelligence should be able to do it all a lot more cheaply?” How do you answer a question like that and make your value proposition clear? The truth is there are many ways to futureproof your agency, and futureproofing will be a main focus of the chapters that follow.
This book is in three parts. The first will help you better understand the specific roles artificial intelligence plays in PPC advertising, separating fact from fiction and what humans do best from what machines do best. The second part focuses on essential roles or personas — creative, client-facing, and technical — PPC professionals will continue to play, since AI can’t. One of these roles will probably appeal to you more than the others, and that is where you will want to focus your energy. The third part deals with positioning and futureproofing your PPC agency in “an AI world.” How do you frame or reframe your value proposition to clients eager to hop on the latest bandwagon, even if they have no idea where it’s going?
The first part is background information about what PPC machine learning is. The second and third parts contain more actionable information about positioning or repositioning yourself or your agency in a machine learning world. Some of you may wish to go straight to part two or three, where the rubber meets the road. The part one background, however, will probably help you, the PPC professional, better understand the often-confusing trends dominating the industry today.
AdWords created the PPC industry. Rebranded as Google Ads, it is still by far the most powerful and influential force in the space. Because I was fortunate enough to be a key member of the AdWords team for many years, I’d like to start by giving you a front-row seat for a review of how machine learning was introduced and then became an evermore-integral part of PPC advertising. This will contextualize a concept — AI — that’s often used vaguely and out of context. Learning more about the “AI-first” PPC world, you’ll be better able to position yourself to become one of its most productive citizens.
***
To keep reading, pick up your copy of Digital Marketing in an AI World: Futureproofing Your PPC Agency by Frederick Vallaeys.
