With automation everywhere, the push to allow Google even more control over our marketing campaigns is not lost on us digital advertisers. The rise of machine learning has meant that Google’s algorithms have become smarter, and with it the number of automations available to us has risen too. One of these options available to us is the use of auto-apply Google Ads recommendations.
For the uninitiated, Google recommendations are a set of guides in the ad platform that suggest how to optimise your campaigns. Should these be applied across all your Google ad campaigns as a default? Or should that fire-and-forget approach make you nervous?
Google’s recommendation is to apply a minimum of 10 auto-apply recommendations to every campaign, and there are definitely benefits in doing so. An automated system will inevitably be more reactive to changes in CPA, or have higher visibility of ongoing changes in search volume. However, we should be wary of the balance between automation and advertiser control. Google may invest heavily into making the smartest ad platform around. But it is the marketers who are the experts in understanding what we want our campaigns to do. And it’s up to us to understand whether a certain recommendation is best for our clients – or best for Google.
After all, it’s simply not in Google’s interest to tell you to put less budget into your campaign. But is that the right thing for your client?
We might be living in the early days of an AI revolution. But the rules of automation have not really changed since the days of the Spinning Jenny. Let the machine do its thing to increase productivity, but never take your eye off what it’s doing.
With that time-honoured principle in mind, here are the top three most dangerous auto-apply recommendations. And why you should think twice before enabling them.
1. Remove redundant keywords
This recommendation has had a recent update, so don’t be caught out if it’s one you have applied already. Previously, enabling this allowed Google to remove redundant keywords of the same match type within the same ad group. Sounds sensible right?
However, from January 19th this year, the automatic policy is now able to remove keywords across all three match types. This involves not only changing keyword strategy that marketers may have been perfecting for years. It also makes us wonder what other changes Google may make to recommendations that we have already applied. Maybe not so sensible.
2. Improve responsive search ads
Allowing Google to improve your responsive search ads (RSAs) means it will use existing content to make changes to your ad copy. This includes information from the landing page, or from assets in the same ad group. This sounds more than a little risky to me.
There are two main issues. First, if you need to get client approval on your copy in order to maintain a certain style, or to adhere to brand safety guidelines, it’s probably not the best idea to allow Google to make direct changes to your brand’s tone of voice. Let’s avoid those unhappy stakeholders.
Secondly, although the changes made are based on existing copy, they’re often not relevant in the context of an ad. For example, we saw the addition of the decidedly unsexy headline ‘Programmes Archive’ into one of our Higher Education clients’ ads, which had no relevance whatsoever to the University course we were promoting. Unsurprisingly, performance for that ad group was well below the norm.
3. Add new keywords
This may sound like an easy win. Why wouldn’t you want to target people searching for similar keywords, therefore expanding your audience and overall reach from your campaigns? Well, to put it simply, you know more about your targeting intentions than Google does. (Or at least, you should.)
Take another example from one of our clients. We offer purely online master’s degrees, and therefore want to put our budget to best use on the most relevant searches we are seeing. Yet we are still being recommended more generic ‘course’ keywords, and generally irrelevant searches. I would stick to manually reviewing the recommendations tab for this one, but take it with a pinch of salt.
Google Ads is an ever-changing platform. So what’s true today may be hopelessly out of date in six months. It stands to reason that Google’s recommendations will evolve and improve over time, and auto-applying all of them should become less risky. But my recommendation is to experiment where you can, and see which ones work for you.
However, when testing I would strongly suggest keeping a close eye on your ads account, and treading with caution…