Coping with the uncertain future of ad automation

The past is not always a good guide to the future, says Niki Grant, Head of Search, The Kite Factory; but the better informed you are, the more able you will be to adapt…

I’m a big fan of Google’s product update events: over my almost-decade-and-a-half career focused largely on paid search, I have come to look forward to these events with enthusiasm akin to that of a child on Christmas Eve. Historically, refrains such as ‘the year of the mobile’ have provided context to the approach or direction upon which Google has embarked. Rising mobile usage figures and developments in the technology brought about Enhanced Campaigns: an updated campaign type which served ads across multiple devices, capable of making the necessary choices and adjustments to serve all devices from a single campaign.

During Google’s February UK Search Summit and March’s EMEA H1 Product Kickoff, however, the usually well-researched, coherent refrain became ‘you won’t know until you test it’. Aside from the usual consumer market or behavioural data to support the developments, the updates themselves are indicative of increasingly unavoidable automations, and a decrease in control for Google advertisers.

There have been many incremental updates to Google Ads tools and products to aid automation efforts; here are a few you might be familiar with:

Creative Formats: The ‘original’ ad format is now known as ‘Standard Text Ads’ [STAs] (containing one headline and two description lines allowing 25 characters and 2×35 characters respectively). In 2016, Enhanced Text Ads [ETAs] testing began and STAs were retired in favour of the longer ad format. Now ETAs are for the chop, with Google’s Responsive Search Ads [RSAs] soon to become the only Search editable ad format (from 30 June 2022).

This wouldn’t be such a concern if RSA performance was on par with that of ETAs, and if Google’s RSA rating system was more robust. RSA ratings often report ‘poor quality’ against even the most keyword-dense of assets.

Bidding Capabilities: Enhanced CPCs were introduced in 2010 to allow Google to adjust a bid by up to 30 percent to receive more conversions. This was the first step away from manual bidding and has since blossomed into a range of bidding strategies, both via Google Ads ‘Smart Bidding’ and through broader GMP tools such as SA360. Whilst I understand Google has a world’s worth of data to back up the decisions their bid strategies are making, without possible transparency in how this world of data is referenced, it is rather a leap of faith to entrust Google with these micro financial decisions.

Recent updates to bidding strategies include the removal of the conversion floor (minimum 15 conversions required over the past 30 days) for campaigns requiring ‘target ROAS’ bidding. Whether born from a desire to force-feed the algorithm more data for more robust decisions, or to demonstrate that the target ROAS strategy can flourish even without a large seed data set, this lowering of the barrier to entry for more commercially focused strategies is sure to be broadly welcomed.

Match Type Changes: Follow Broad Match Modifier [BMMs] introduction in the early 2010’s, last year Google announced that BMM would be retired out and that the phrase match function would expand to pick up the slack. This shuffling of match type options seems inconsequential however, at recent events Google have been recommending the use of Broad Match alongside automated bidding and RSAs to improve performance. Broad match is widely regarded to be the sledgehammer of approaches in the sense that it certainly gets the job done but with a fair amount of unnecessary mess. Google’s recent positioning of the problematic match type is that it is an ‘upgrade’, and ‘automatically finds millions of keywords’.

As you may have noticed from the above details, automated functions are no longer just an ‘opt-in’ feature – whether you proactively adopt automation more broadly or not, Google will bake it into the core functions of your campaigns. Here are five tips for brands to help survive and thrive in the uncertain age of automation:

  1. Keep up to date: Do your best to keep up with transition dates for new tools to better prepare for fluctuations in performance and required input. (Make a note of 30 June as the cutoff for new ETA creation!)
  2. Use 70:20:10 to test: Where updates are optional or phased use the 70:20:10 approach to test alternative tactics. Use this ratio to allocate budget across ‘Proven Approach: Refining an Approach: Testing Something Totally New’ to mitigate risk in testing investments.
  3. Nail your conversion tracking: Whilst Google increasingly automates campaign decisions and optimisations, the in-Google conversion tracking will be one of the best data sources for the algorithm to cite and consider (if you don’t have SA360/GA integrations with all the bells and whistles).
  4. Take a holistic view: The Google algorithm takes a huge number of different factors into account when making decisions and if we judge the impact of these large-scale decisions with small-scale performance, we’ll never recognize the benefit. Ensure ‘macro’ performance views are considered in line with your chosen attribution model.
  5. Optimise on-site content: Increased reliance on automated ad copy creation means increased reliance on the content you have available, either as reference to craft that ad copy, or for use on your landing pages. Quality content should stand you in good stead for relevance signals (as ever) as well as effective future dynamic ad creation.

By Niki Grant​, Head of Search, as featured in Performance Marketing World