The ascendance of automation
Over the last few years, we’ve continued to see advertising platforms limit the number of manual controls available to advertisers, instead using machine learning to inform decisions around ad delivery. As an advertiser, this can feel like you have less control over your Google ads budget than you did several years ago, leading some to reject Google’s use of automation.
This rejection is a huge missed opportunity for advertisers. Like it or not, paid search smart automations are the future of online advertising, so we must learn to work with the machines to gain a competitive advantage. In this post, we’ll explore the limitations of Google’s machine learning, as well as how you can make the algorithm work for you to push past these limits and maximize profitability.
What are the limitations of Google’s machine learning?
Since the inception of smart bidding in 2016, Google’s machine learning capabilities have grown exponentially. Smart bidding strategies now take hundreds of thousands of context signals into account when setting bids in the ad auction, all in real-time.
However, one of the main limitations of Google’s machine learning is that the algorithm can only utilize the performance signals that it has access to. This can present a big challenge for lead generation clients in particular, as Google doesn’t have oversight over whether the prospect actually led to a sale. A similar limitation applies to e-commerce accounts, as Google’s understanding of your business doesn’t extend to profit margins or item returns.
In order to work around this, we must bring our unique understanding of our business to the table, providing Google with actionable data that the algorithm can use to make more informed bidding decisions. Fortunately for us, we have several tools at our disposal.
1. Seasonality adjustments
One powerful tool to augment our smart bidding strategy is seasonality adjustments. This advanced control enables us to inform our bid strategy when we’re expecting a significant uplift in conversion rate. This allows us to further Google’s understanding of events, such as flash sales or product promotions that it would otherwise not be able to predict.
Seasonality adjustments also offer a degree of granularity during setup; you can apply the adjustment during specific dates, to specific campaigns and across selected device types. The controls also allow you to specify how much you expect your conversion rate to increase, so ensure that you conduct thorough analysis beforehand to make sure this is as accurate as possible.
You can find these controls by navigating to the ‘bid strategies’ tab (under tools & settings) and selecting ‘advanced controls’ on the left-hand side.
2. Conversion value rules
The second tool you can use to make Google’s machine learning work for you are conversion value rules. When using a ‘maximise conversion value’ strategy, these rules allow you to adjust conversion values based on actual value to your business, using conditions such as device, location and audience.
This is a really useful way of introducing business insights to your Google ads campaigns. For example, if you know that product returns for your e-commerce business are particularly high within a specific location, you can set a conversion value rule to adjust the value accordingly. Similarly, if you know that the lead-to-sale conversion rate for your lead gen business is much higher when prospects submit an enquiry on a desktop device, you can increase the conversion value so that Google’s machine learning will prioritise desktop auctions.
Conversion value rules even allow you to set multiple conditions at once, providing a high level of granularity when applying these rules. If you’d like to give this a try, you can find the controls by navigating to the ‘conversions’ tab (under tools & settings) and selecting ‘value rules’ on the left-hand side.
3. Uploading offline conversions
The final method of assisting your smart bidding strategies is to simply provide the algorithm with more conversion data to optimise towards. Google ads allows you to upload your offline conversion data when these conversions are linked to a Google click ID (GCLID), meaning that your bid strategies can optimize toward actions that take place outside of your website.
The most obvious application is for lead gen businesses, as this feature allows you to upload closed sales data. This means that the algorithm can now work towards driving high-quality leads rather than delivering a high volume of low-quality enquiries.
There is a little bit of setup time required to make the most of this feature, such as having auto-tagging enabled and being able to capture and store GCLID data, but having your smart bidding strategy deliver real, tangible results is more than worth the investment.
You can start uploading your offline conversion data by navigating to the ‘conversions’ tab and selecting ‘uploads’ on the left-hand side. You can even schedule this data to upload automatically by selecting ‘schedules’ at the top of the page and selecting a source from the drop-down menu.
Take back control
In this article, we’ve covered several different tools you can use to take back control of your smart bidding strategies. The key takeaway from our discussion is that, although machine learning is the future, our unique understanding of our business is ultimately our greatest asset to make our campaigns more profitable.
Instead of rejecting ad automation, we must instead provide the platform with the context it needs to make informed decisions on our behalf, making sure your budget is working for you and not for the platform.
Partner with the PPC experts
By partnering with a performance-driven digital marketing agency such as Circulate, you can ensure your PPC efforts work hard for you, ultimately, helping your paid advertising budgets go much further.
We offer a range of specialist digital services that are here to help your business grow, so contact us today or book a free strategy call to see how our friendly team of experts can help your business grow.
