Skip to content Skip to footer


Adapting to Google's Machine-Driven Ad Revolution

by Michael Patten, Digital Marketing Strategist, 13 June 2019

Read Time: 5 minutes

It’s funny to think back to around eighteen months ago when the PPC industry was publishing volumes of paranoia-infused articles focusing on the notion that machine-learning assisted bidding could render human-led account management obsolete.

In reality, the many and varied ‘smart solutions’ Google has incorporated into the platform have in fact shifted the work away from repetitive and manual tasks, replacing them with a new challenge – keeping the automation performing at its full potential and providing the wider context that only a human can interpret.

Google cartoon of Machines and Humans happily working together

Source: Google’s Academy for Ads

I still keep my tin foil hat half-placed on my head with regards to the theory that if Google controls all of the bidding activity across the platform, it makes it very easy for it to tweak the algorithm to generate the best profit for itself. Let’s face it: Google has always been the real winner, with around 86% of its turnover being generated by advertising activity alone.

But resisting – or even outright denying – the importance of incorporating machine learning is futile. When you really think about it, machine learning is actually a long-standing feature if you consider how optimised ad rotation, dynamic ad extensions and budget flex have functioned for many years.

Changing your strategy

Consider just how many characteristics can be identified about a user (which also allows for bids to be modified and ad exposure altered accordingly). It makes perfect sense that bidding is computerised and the manual ‘broad strokes’ of traditional, human-led bid setting can no longer remain as effective.

Google has also provided a decent variety of options in terms of what approach to Smart bidding you can incorporate, with automated strategies suited to differing situations that appear to be constantly updated and evolved from Google’s side.

A big sticking point used to be that strategies aimed at growing or refining conversion volume were not best suited to smaller businesses and accounts. Now, the main strategies are able to function without needing a minimum conversion threshold to be met first, and even aim to make improvements in tougher learning conditions – such as having a restricted budget or low impression share.

Three of the main automated bidding strategies offered by Google: Maximising conversions, target cost per acquisition and target return on ad spend.

Source: Tomaso Uliana - Presentation on Smart Bidding

Although it’s important to note that where these solutions are being constantly changed, it is our performance data that is shaping their success. In effect, account managers are now inadvertent beta-testers for Google.

It’s not hard to understand how such change within a short time frame can be difficult to stomach for some. In order to make the most of this industry-wide transformation, both the marketer and the business needs to adapt their approach, not only to best understand the conditions currently in effect, but also to realise the best return that can be made from all of this new technology.

The beginning of this is in realising that automated bidding not only needs time to learn, but also that it rarely displays the ‘sudden impacts’ that we are used to seeing through manual bidding. Assessing the success of an automated bidding strategy requires more patience, particularly when you consider how the machine learns by observing user journeys. Consideration is needed for the time it takes from a user’s first ad click to when their conversion action occurs.

Graph depicting the period in time between an automated strategy beginning to learn and produce results

A real-life example of a three-month period in which the strategy learns for the first two, and then delivers results in time for the third.

Source: Screenshot taken from the Google Ads interface

A large factor in the effectiveness of an automated strategy also lies in what you feed it. Or, more specifically, providing enough relevant and varied content for use as lines of ad copy or creatives.

It is also very easy to overlook keywords in today’s industry, now that keyword-to-search term matching is possibly the broadest it has ever been. Being proactive at keyword (and negative keyword) research, along with reviewing any dynamic ad content being served for relevancy and clarity is still very much in the hands of the human marketer.

Anthony Chavez on stage at The Google Ads Innovation Keynote

Source: Google Marketing Live 2019 - Ads Innovation Keynote

What’s next?

2019 is already (and predictably) shaping up to be a big year for the PPC industry. It’s no secret that the days of manual bidding and static copy are numbered, with Google recently announcing that even more automated features are coming to the platform later this year. Adapting the approaches, processes and mind sets for today is the key to achieving tomorrow’s marketing goals.


Insights straight to your inbox

Sign up now to be the first to know when our experts publish a new article.

Your information helps us monitor and improve our content, we will never share it with anyone else. We only send emails we think you’ll find valuable – you won’t hear from us more than once a week. You can unsubscribe any time.