Machine learning has been around in search marketing for over a couple of years now. Typically, the adoption of these new technologies leads to uncertainty and friction. Many advertisers therefore try to force their traditional digital ways of working and refuse to make tough decisions such as embedding automated tools for search marketing. This outdated mindset is hindering them from making a full digital transformation.
Businesses should embrace automation in search marketing and develop the skills to understand it, albeit with a critical mindset. Now is the time to reset, pivot, and think big to reevaluate your search marketing automation to match digital expectations. This change in mindset will enable companies to climb the digital maturity curve and consequently achieve incremental performance.
Complex online ecosystem
Over the years, a paradigm shift took place, resulting in a complex online ecosystem. Consumer behavior reshaped the search ecosystem with the merge of mobile devices and more complex signals to take into account. Consumers are now performing online research through a vast range of connected devices, with the online and offline world completely being intertwined. The COVID-19 pandemic has caused a decade’s worth of digital disruption in just a matter of months. This has resulted in digital shopping behaviors and expectations being more sophisticated than ever and making it merely impossible to manually manage all these signals.
Given these recent market and consumer disruptions, Google has incorporated more and more automated solutions into products like smart bidding, smart campaign types and smart creatives.
How automation is shaping Google Ads
Several years ago, the daily tasks of a digital marketer consisted of performing thorough keyword research, manually creating tailored ad copy, setting manual bids and creating overly complex account structures.
This practice was common back then since relevancy was more than ever important. We needed tight control to maintain efficiency because we were manually monitoring every aspect. However, this manual way of working was extremely time-consuming, leaving little headroom for strategic tasks.
Machine learning opened up the door to identify insights, detect patterns and identify opportunities. Algorithms have evolved throughout the years making them more robust and performant to predict and act upon these signals. When smart bidding was first introduced over 4 years ago, technologies weren’t as powerful as they are now.
Marketers were reluctant to use them back then since they generated rather poor results. Furthermore, many weren’t fully trained to use these automated solutions properly as well.
When smart bidding was first introduced over 4 years ago, technologies weren’t as powerful as they are now.
Since mid-2019, automation algorithms have extremely evolved and are now more robust than ever. Backed with matured automated solutions and better training, we have used these solutions ever since at The Reference. We have many colleagues that have adorned the label “digital marketing consultant” here at The Reference, who only now feel like they are finally able to become a marketer.
But what are the reasons a lot of advertisers and consultants have not yet embedded automation in their daily operational tasks as of today?
The lack of control and transparency in search marketing as a result of these automated solutions is often touted as a key argument.
Automation and lack of transparency?
What are the different aspects that are being automated in search marketing and what is their impact on transparency and control?
Automated bidding in Google Ads uses machine learning to set the right bid for each and every auction and uses historical data to predict outcomes to help you achieve your goals.
Automatic bidding has been a major paradigm shift in search marketing and is probably the most discussed item with mixed opinions. Letting Google set your bids scares a lot of people and we can understand why.
Should you therefore not adopt this technology? Absolutely not.
A savvy marketer will use this technology in a layered approach starting with experiments, evaluations and using the outcomes for other campaigns. When the experiments have guided you towards your ideal way of working, you can start transitioning towards a full implementation.
Ad copy management
Google has launched automated solutions to create semi-automated ads, called responsive text ads.
They are partially automated since they still require our textual input, but ad variations are optimized based on performance. You can use Google’s automated “Ad Strength score” to know how to improve your ad quality.
If you are ready to give away more control and have the machine help you with your ad copy variations, you could opt in to the automated ad suggestions. But again, this is a personal choice and you can even opt out of this feature if you prefer to write your own compelling ads.
All keywords are now semantic instead of syntactic, which means advertisers can drive valuable results for their businesses without having to perform detailed keyword analysis. We used to fully rely on search term insights to evaluate the performances of our campaigns. However, Google recently communicated they would reduce the visibility of search terms shown in reports. This loss of transparency has quickly resulted in a lot of friction from advertisers. And we do understand that.
However, we would rather advise to focus on the greater picture and review search patterns rather than mining low volume search terms.
There is a vast range of smart campaign solutions available, ranging from automated display campaigns to automated shopping campaigns.
We have seen amazing results with smart shopping campaigns for example and would have to agree the results outperform traditional shopping campaigns.
You are free to choose if you want to create smart campaigns, but we would definitely recommend you to at least give it a try.
Mid-2020, Google has launched recommendations to simplify Google Ads structures and consolidate traffic in centralized ad groups in order to reach sufficient volumes and let Google’s algorithms work at their best.
This update definitely generated a lot of criticism, especially from consultants being in the field for a couple of years now. We do have to admit that at first, we were some of the first people to question this approach.
Automation helped us stay focused on what really matters - growing and achieving your KPIs and doing what we humans are good at: creating strategies.
However, we do have to agree with the approach to simplify current structures and definitely getting rid of hypersegmented account structures (based on audiences, devices, keyword types etc). Now is the time to start revising the current account structure, revise the current keyword selection and eliminate synonyms and plurals, for example.
Furthermore, We’re also convinced that “portfolio bidding strategies” are a great solution to start the transition towards a simplified structure for those of you that don't want to make big changes at first or even for grouping campaigns with lower volumes.
Preparing for the road ahead
It is clear that automation has become an imminent part of search marketing, and even in marketing in general. Automated search marketing systems have truly leapfrogged manual work and it is no fantasy to assume this trend to continue in the years to come.
From our experience, we would have to agree that the transition towards automated solutions required a bold mindset. As an agency. As a consultant. As an advertiser.
It required patience in being properly trained in how to manage these solutions. It often takes time to unlock a tool’s benefits and yes, adopting new tools will not work in every instance. It’s not going to be perfect on day one - but we know very few things that are.
Most importantly it has freed up time for added value work. Automation helped us stay focused on what really matters - growing and achieving your KPIs and doing what we humans are good at: creating strategies. We are most definitely not outperforming machines on processing all available data points in the current online ecosystem.
Yes, we had to let go of a part of the control we used to have, however always with a very critical mindset and not blindly following all advice. It’s all about finding the right balance between control and what truly matters. The balance between automation and autonomy.
Don’t be afraid to loosen your grip on manual optimizations and let automation take the driver’s seat. Let your expertise and common sense be your passenger.