.png)
Who we are
Who we are
BLOG POSTS
BLOG POSTS
BLOG POSTS
BLOG POSTS
The implications of AI driven analytics
Over the years, Google have implemented many AI powered tools that aim to ease the efforts of campaign building, monitoring and optimising for those in marketing. In this blog, I will talk about examples of changes as well as my perspective on how this will affect the future of those working with the software on a day-to-day basis.
Automated bidding – The learning phase
There are many AI driven tools within the platform. One example of this is smart bidding, an automated bidding tool within Google Ads. Google uses a set of automated bidding strategies to optimise towards the right conversion goal. For example, maximise conversions, enhanced CPC and Target ROAS. The primary implication that many find with this AI tool is the learning phase. As the strategy uses algorithms, there is an adjustment period in which it has to learn what works best for the campaign. In this period it uses different signals such as CTR, conversion rate and engagement metrics to test what work best with your audience, depending on what your set goal is. It is likely that this may use budget excessively, increase your CPC, all whilst not necessarily generating more clicks or conversions. In my experience this usually lasts two or so weeks before performance stabilizes.
Ad ranking
Although Google does mainly base which Ads perform well based on data collected, it is important to be aware of Ad ranking. This is a set of values that determine how often your Ad is shown. Factors include image formatting, the budget set being too low or not using a strong number of keywords within your Ads. If your ad is ranked ‘poor’ it is likely to be served less than an Ad that is served ‘excellent’. It can be common for your Ads to rank poorly during the learning phase whilst the algorithm learns what works best.
Overall opinion on smart bidding
The process is often still very manual at the start as it is important to monitor the campaign’s progress throughout the learning phase. However, the process can be sped up if historical data is implemented meaning the campaign has substantial data behind it so that it does not need to use algorithms to learn to the same extent if it was a brand-new campaign. Bottom line, once it has gone through the learning phase it does seem to be more time efficient than manual bidding, but sometimes a bid strategy needs to be changed and this does take a bit of trial and error to optimise.
My experience with AI within Google Ads
One-word, PMAX. Love it or hate it, it is a tool those in paid search are well-versed in. PMAX can be run across many channels including display, search and discovery. The tool is an automated means of creating and monitoring a campaign. It works by providing the performance max campaign with assets and setting the campaign to optimise towards a conversion goal. It then uses real time ‘smart bidding’ to heavily drive the campaign.
My experience with this is that PMAX loves to spend money, specifically on new potential customers. It is also a good idea to bid towards maximum conversions. This can most definitely be a positive thing as it widens your audience to those with related interests you may not have been aware of before. However, it allows little insight into the conversion journey and tracking. I have found that for some clients it works and for others not so much. It usually has a longer learning period which takes close monitoring and patience, but it is good insight into what the future of marketing holds.
Final thoughts
The development of the industry over the past few years has been revolutionary and it will never stop expanding. Within digital marketing it is important to keep adapting and be conscious in our efforts to do so, as technology will develop with or without us. Complete transparency with clients and taking the time to learn how to work with new tools, could make for a smoother transition. The upshot of it is that Google uses these tools in conjunction with human efforts to maintain it. It is still part of our role to provide Google with the appropriate assets and for the moment AI does not have the capability to do everything from start to finish as humans pick up things that algorithms can’t. I believe transparency and mutual understanding with the client, optimising towards clients’ needs and goals should still be at the forefront for the future. Being absolutely clear of the objectives and what it is we are optimising towards creates the stepping stone for how to get there.
Written by Eve Francis
References-
https://www.wordstream.com/blog/ws/2018/12/19/google-ads-automated-bidding
https://www.sfdigital.co.uk/blog/what-to-do-with-the-google-ads-learning-phase/
https://decocre.com/blog/google-ads-learning-phase/
https://support.google.com/google-ads/answer/156066?hl=en-GB
