How to find the balance between creativity and automation in PPC
Here's how you can use the potent combination of human input and machine learning to create winning PPC campaigns.
Day 2 of SMX Advanced 2022 has kicked off and this morning’s keynote by Brad Geddes was all about leveraging automation in your ad campaigns.
Geddes is a PPC expert and co-founder of Adalysis, and the author of “Advanced Google AdWords,” the most advanced book ever written about Google’s advertising program. Geddes has worked with many of the world’s leading companies in managing and perfecting their PPC management and workflows.
Where is the balance?
In his keynote, Geddes went into great detail on the difference between humans and machines, what their strengths are, and how they can work together to create a winning ad campaign.
“Google and Microsoft aren’t taking away your control,” Geddes said. “They’re giving you more management options. You don’t have to fight the machine, but your job is to find the balance.”
So what is the difference?
Realizing what humans and machines do well and playing to those strengths is key. Humans are really good at:
- Reacting quickly to market changes
- Auditing the machine
Machines, on the other hand, are really good at:
- Statistical significance
- Finding lookalike audiences
- Inputting repeatable data like reports
- Conversations from human-driven inputs like chatbots
Referencing recent PPC survey results, Geddes reminded us that ad managers are happy with the results when it comes to scripts and bidding. Managers have neutral feelings when it comes to RSAs, data-driven attribution, and local campaigns.
Not surprisingly, managers are generally unhappy with automation surrounding discovery campaigns, the insights tab and auto-applied recommendations.
All in all, machines are really good with numbers, but not with insights and intent. That’s where humans come in.
Is anyone surprised? I didn’t think so.
What do we want? Balance of data insights combined with a machine’s ability to crunch numbers and make predictable outcomes!
When do we want it? Now!
Geddes described guardrails as boundaries advertisers put around automation.
Google doesn’t always get it right and adding guardrails helps advertisers “leverage automation within our own framework of tolerance and profitability.” Geddes goes into depth on how to use these to fine-tune your campaigns.
A few of the most common guardrails are:
- Negative keywords
- N-gram analysis
- Negative audiences
Let’s talk strategy
One of the most important takeaways in Geddes’ keynote is how to address key parts of your funnel with proper messaging and measurement.
How is the message changed based on previous user interaction? Is there cross-channel integration to consider?
But how do you create a plan around keywords, audiences and landing page experience?
The right campaign builds for almost any account size & type
Consider your campaign type, budget and strategic goals. Then think about your strategy.
Geddes explained the exact campaign builds he uses for his own accounts which include:
- Search or display
- Custom audiences
- Exact or phrase match – or broad match with a lot of data
- Lightly pinned RSAs
Complications with strategy or campaign type (such as smart shopping or video) require additional considerations. However, as Geddes explained, this campaign build can work for everyone and uses several types of automation at the same time.
When should we override the automation
Geddes said that in reality, especially with RSAs, we have more control than we think. We aren’t forced to use machine learning. Account success is based upon KPIs, not on how much automation you’re using.
Don’t undervalue humans
Sometimes you want full automation, but Geddes laid out some considerations before diving in.
Watch: Finding the balance between creativity and automation
Below is the complete video of Geddes’ SMX Advanced keynote.
New on Search Engine Land