Q & A With Patrick Gilbert on Digital Advertising in the Age of Automation

This conversation initially took place in Julie Bacchini’s PPC chat twitter sessions. Patrick Gilbert joined her to discuss automation in digital advertising. They explored common misconceptions around automation, and the role PPC managers play in the age of automation. 

Q1: How much automation are you currently using in your accounts? Does it vary by platform?

Across our Google Ads MCC, 93% of our client spend is on smart bidding. The exceptions are some branded campaigns and display campaigns. Facebook Ads has always been 100% automated, although most people don’t tend to think of it that way.

Q2: What is your general attitude regarding automation in PPC? Has it changed in the past year?

In 2017 we saw the writing on the wall and embraced automation. Our overall attitude has not changed since then… We assume, first, that automation is the right solution; only then can we look for exceptions to the rule.

Too many people blame Google’s automation for their failed ad campaigns, and instead need to take responsibility for their work. More people need to be asking themselves, “How can I improve the way I use this technology to drive performance?”

If our automated campaigns fail, we don’t blame the system for “burning through our budget.” Perhaps we’re not feeding the system the proper data? Maybe it’s a landing page or a product issue?

Automation is incredibly complex. You can’t just start using it without investing in your own education. It takes time. Since becoming a pro-automation agency, we’ve 3x’ed in size & our client’s campaigns are more profitable than ever. This is not a coincidence.‍

Q3: Why do you think the industry has been so slow to embrace automation?‍

People don’t like change. And few will admit this, but many are afraid that automation will make them irrelevant. I used to feel the same way… I was anti-automation for a very long time, simply because I was insecure about the future of my career

In 2015, most of my day was spent building SKAG campaigns, making segmented bid adjustments, and finding negative keywords in a search term report. None of those things matter anymore. My job, as it existed then, was technically replaced by automation.

My first experiences with automation were disastrous. I was pressured into it by Google or a client, without a true understanding of how the tech worked, and we burned through our budgets with terrible results. We lost clients from this.

It’s obvious, then, that we’d be hesitant to welcome automation with open arms when it always seemed to lose us and our clients’ money! Over time, we realized there would be a learning curve and changed our attitude. We also recognized the inevitability of it all.‍

Q4: What are the common misconceptions around automation?‍

People often try to answer the wrong questions. It’s not about automation vs. manual; it’s not about Maximize Conversions vs. tCPA; it’s not about RSAs vs. ETAs… It’s about the data! It’s about whether you’re setting your algorithms up for success! 

Without the right data, you’re either miscommunicating your goals to the algorithm or you’re sending it on a wild goose chase. You will never truly understand all that is happening on the backend. So instead, focus your attention on solving real business problems. 

For example, stop testing Smart Shopping vs. Standard. Just turn off Std. or leave it alone, I don’t care. Your attention should be on how to best leverage the New Customer Acquisition feature. How can you train Google to understand the value of a new customer? 

Q5: Are there scenarios where automation struggles in your experience? For example, low volume accounts?

Automation always works better with more data. Larger accounts will have an easier time finding success. That doesn’t mean it’s impossible for smaller accounts… it just means it’s more of a challenge. You have to get creative w/ the data that you’re feeding it.

My recent presentation at SMX discussed how to control the learning environment to expedite the algorithmic learning process. This is key for low-volume accounts. A recording of the video can be found here: bit.ly/3o6X5WY 

Also, consider this – If you’re in an industry where you are willing to use automation, but your competitors are not (due to low volume, etc.), it presents a massive opportunity for you to capture the highest intent auctions that others are ignoring!

Q6: What is the role of the PPC manager or freelancer in the age of automation? Are they still relevant?

Automation will change your job, but it will not replace you. In fact, the opportunity to provide value to your clients is now greater than it has ever been. The PPC Manager just needs to be willing to learn different skills.

Today, we spend much less time working inside of accounts, and much more time developing strategy. I like to say that we don’t manage campaigns for our clients; we find creative solutions to their complex problems (via digital advertising tools).

Today, it requires a higher level of business acumen. PPC Managers need a higher sense of data literacy than ever before; they need to understand good creative, good copy, good product design, and how to tie all of this together in a simple advertisement.

Pre-automation PPC was basically a sales job. The PPC Manager was responsible for connecting the dots between a bottom-funnel search for a product, and a landing page that sold that product. Today, it actually requires real marketing! It’s so much better! 

Q7: What does a modern, pro-automation PPC agency look like? Are you currently moving toward becoming a more pro-automation agency/consultant? If not, why not? How close are you currently to being a pro-automation PPC pro or agency?

We used to be a team full of account managers. Everybody did everything: conduct research, build campaigns, manage data, build reporting dashboards, manage client relationships, write ad copy, create Facebook graphics. In hindsight, it was absolute chaos.

Our original Account Management team now has two divisions: Client Services and Strategy. CS manages the client relationship and campaign execution; Strategy helps solve the larger marketing problems. Both are responsible for campaign performance… 

We’ve also been building out an Analytics team that can do the real data analysis that I’d never been able to do. Things like linear optimization and complex correlation studies using python… Strategy is becoming more data-driven…

For example, our analytics team is diving into complicated attribution questions, or calculating customer LTV, or finding correlations between a TV ad buy and on-site conversion rate. It’s really exciting. These are things we NEVER did before automation...