PPC Automation: The Challenges of Algorithm-driven Google Ads

PPC or Pay-Per-Click is one of the most lucrative digital advertising channels, with Google Ads accounting for nearly $120 billion (about 85% of the company’s global revenues). Technology developers as well as publishing platforms have recently started creating new A.I. and other algorithm-based solutions to automate virtually all aspects of the PPC functions and operations and recently a Quora user asked Chris from Worldacquire whether this would mean the end for PPC service providers such as PPC agencies.

 

Chris, who previously managed PPC and digital marketing client-side at American Express and viagogo, and extensively worked with and designed PPC automation solutions, answered the following:

 

“There are three variables to consider when answering this question:

 

1. The reliability and effectiveness of automation
2. The unbiasedness of automation solutions
3. The adaptability of PPC agencies

 

Many aspects of typical PPC work can be automated in multiple ways, from campaign management to bidding strategies.

 

But how to determine which solutions work for you?

And will they always work 100% in your favour?

Are the algorithms and automation solutions released by publishers like Google or third party platforms like Marin and Kenshoo completely unbiased, or will they “automatically” prioritize increasing their own margins?

 

My personal experience (client-side PPC/Performance Marketing/Biddable Lead at American Express and viagogo, consulting for Rolex and TAG Heuer, and through Worldacquire) has shown that PPC automation comes with:

 

Many errors and bugs, determined not only by discrepancies in the automation code/instructions, but also in the statistical goodness of fit of the models created to calculate the changes required to reach the set goals. You might need to test and/or concurrently use multiple automation solutions in order to find the right answer for you, and this will also change depending on the specific product and product cycle. Perhaps you can build an automation solution to automate this verification and selection process, too, but it’s going to take a while and a lot of data to make reliable decisions.

 

Bid changes that somehow appear to be more favourable for the publisher or third party platform’s own margins instead of yours. How can you prove and ensure that this doesn’t happen? How can you monitor and verify tiny and apparently random changes that could actually have a big negative impact on your end of month ROI/ROAS and Net Profit?

 

Short-lived results. Assume all the above is solved, works perfectly and you (or your A.I. assistant) build great processes and have found the right algorithm. How likely is this to last for more than a few months? Can it also automatically adapt to the endless updates of publishers’ definitions of Quality Score, the ever-changing industry CPC thresholds, product changes, market/customer behaviour shifts, new competitors, etc? Can it proactively come up with completely unimaginable and innovative solutions on ad copy creation and ad design that will out-Google all high quality CTR standards?

 

Whatever PPC automation solution(s) you adopt, you (or someone else like Worldacquire) will need to constantly ask the above questions and assess their strategic and financial impact on the marketing department and the company’s bottom line.

 

Many PPC agencies are traditionally used to focusing only on operating PPC services and fitting PPC services into template strategies. They will need to learn how to tweak PPC services and shape the entire strategy by taking into account PPC (and other areas of) automation.

 

Conclusion: automation won’t make PPC agencies redundant, but it will force many of them to change the scope of their work, to embrace, understand and properly adopt automation and A.I. for their clients.

 

A fair amount of PPC agencies and consultancies around the world has already started adapting years ago. Worldacquire is one of them.