7 ways to kickstart machine learning for PPC in 2019

7 Ways to Kickstart Your Machine Learning PPC Strategy in 2019

Adopt these innovative strategies to increase conversions and improve ROI.

Remember The Jetsons from the 1960’s? Who could forget the images of flying aero-cars and robotic maids and other whimsical inventions, which all seemed like child’s play at the time? But just like George, Jane, Judy & Elroy, families today can talk to their houses to request food delivery and weather updates or schedule meetings. Uber has even announced a flying taxi that is projected to start service by 2023. These types of rapid changes are precisely the reason why digital marketers and agencies must sit up and take notice – instead of “flying by the seat of their pants.”

While machine learning is nothing new, the rapid pace of change by which it has made inroads into all areas of our homes, lives, and businesses in recent years is nothing short of staggering. Everything from AI assistants like Siri or Alexa, GPS direction finding, language translation, smart devices, you name it, the technology is becoming increasingly ubiquitous.

There are some exciting evidences of machine learning technologies in marketing today, such as chatbots, more intelligent email campaigns, and even AI that will produce content for you. But the trend that’s been catching the most attention over the past year is machine learning applied to pay-per-click (PPC). Some case studies have shown clear evidence of lower CPA, lower CPC, higher conversions, and better overall budget attainment through the use of AI algorithms.

Since the technology is constantly changing, now is definitely time to get onboard with the latest tools, technologies, and strategies that will set you apart from your competition. As Jeff Baum over at Hanapin Marketing has well put it, stakeholders simply must get used to this new normal; they must adapt or be left behind.

This article doesn’t attempt to theorize or speculate about the status of machine-based PPC. We take for granted the explosive nature of the industry. Instead, the goal here is to provide a quick hit list of 7 actionable machine learning strategies that will refine your PPC approach, optimise your Google Ads conversions, and improve overall ROI for your business and your clients.

1/. Identify Your Desired Outcome

Like anything else in digital marketing, proper planning and strategy is critical to success when it comes to machine learning. Automation shouldn’t be thought of as the “easy button” for digital marketing, but rather is meant to free you up to manage the higher order tasks. While machines are great at manual jobs, they cannot (at least not yet!) identify all of the KPIs that you should be monitoring. So, to kick things off right, make sure you’re tracking everything from CTRs to CPAs to CPCs. Keep asking yourself what data will help you understand the right types of customer behaviour. What metrics will you need to identify progress towards more conversions?

2/. Focus On Your Best Metric First

There are many important metrics that could be singled out for attention in any PPC campaign. Bringing on new machine learning tools and strategies can present a significant enough learning curve that it may be wise to selectively choose one or two key KPIs to focus on initially.

Once you feel comfortable with getting better results on your CTR, for instance, then migrate to the next KPI. This same approach should apply to your clients. As one analyst writes, “Stick to what clearly indicates progress according to your clients’ standards and don’t overload them with extra KPIs just to look good – less is more when it comes to client reporting.

3/. Use Scripts to Automate the Most Manual Tasks

Getting started with scripts is a good introduction to machine learning in general. There are tons of scripts available that can be copied and pasted into Google Ads to optimise your workflow or increase your PPC performance. But it’s easy to feel overwhelmed about where to start. Fortunately, Google offers a fairly straightforward guide to installing and even writing your own scripts. Once you get comfortable, there are plenty of examples to experiment with. More advanced users can migrate to the Cloud Machine Learning Engine.

The key here is to get started: once you begin using AI algorithms to automate tasks for your business, you will obtain a significant competitive advantage in the market!

4/. Leverage Responsive Search Ads

With AdWords rebranding in June 2018, came several new feature sets in the Google interface designed to simplify, automate, and optimise the PPC campaign setup and management process. Among these changes was Responsive Search Ads. When creating ads users now have the option to setup multiple headlines and descriptions. These are adaptive and adjust content to different combinations of text in order to discover the most relevant ads for your audience. This is essentially the same as an autonomous A/B testing feature, and shows how machine learning can simplify this otherwise manual task.

Leverage Smart Bidding for PPC Optimisation

5/. Ease Into Smart Bidding

Smart Bidding is a feature that combines machine learning and contextual signals such as geolocation, time of day, ad creative, and user device to optimise bids at auction level and help users find the best conversion opportunity. There are plenty of solid reasons why someone would incorporate smart bidding into their PPC strategy. But it’s also a strategy that is reserved for more advanced users who understand precisely how these signals align with their marketing goals.

Smart Bidding is not an exact science and may not work ideally with already underperforming campaigns or ones that are not properly with company marketing goals. Use this strategy cautiously until you’re comfortable with the Google Ads ecosystem and supporting machine learning tools.

6/. Keep Expectations in Check

PPC is more complex than ever and requires a keen combination of technological acuity, business strategy, and management insights. While machine learning is a powerful enhancement that allows for the collection and analysis of large amounts of data in ways that humans can never do, it’s not a panacea for all business problems.

Many stakeholders even may be suspicious of machine learning. They need to be taught that it is an important complement to human efforts, not a replacement. In fact, over-reliance on machine learning can have negative effects. Hanapin Marketing says it best on this point: “In short, deploying machines without proper direction will lead to failed performance, wasted budget, and unhappy stakeholders.”

7/. Measure & Adapt For Continuous Improvement

The irony perhaps of machine learning is even though it introduces a time-saving element on manual tasks, it can also create other sets of challenges that may be just as time-consuming to resolve. Integrating the technology into your business without the right processes to support it can be a recipe for failure. The key to success is to consider carefully how you will measure and monitor performance as you benefit from the machine learning technologies. If you’re not seeing improvements, for example, in the Quality Score or if your CTR is below a certain desired KPI, then it’s important to decrease bids and pause ads and monitor the automation until you start to develop fresh patterns and new insights.

What’s Next?

Machine learning has made our lives immensely more streamlined and productive. And though the technology is not perfect (think facial detection backfires like your dog’s face in place of your own), in spite of those occasional false positives things are improving drastically by the day.

7 Ways to Kickstart Machine Learning for Your PPC in 2019Yes, ML has also created a plethora of exciting opportunities and possibilities for PPC strategists today. The wide range of tools and technologies now available on the market can help practitioners save time and help them optimise, analyse, and iterate more sophisticated campaigns than ever thought possible. Smart bidding, responsive search ads, and Google Ads scripts are just some of the many ways in which machine learning is becoming more commonly deployed in PPC strategies.

So, the double-sided question now comes into focus: What will you do with this information and how can you envision leveraging machine learning this year? There is obviously a clear call to “adapt or get left behind,” but what does that mean on the ground for your business?

We have to admit that as cool as all this sounds, there are some tradeoffs that come with these technologies. For example, machine learning is great at processing large amounts of data and automating certain menial tasks, but it can never be a replacement for human-based strategies and processes.

The 7 best practices outlined above recognise that machine learning is best suited when aligned with human efforts to discover the most unique customer behaviours and personas that drive the best conversions. Human-machine interaction is an important attribute of any robust PPC strategy. Getting onboard with this reality is the new normal for today’s progressive minded digital marketer. One writer sums it up best as follows: “The ultimate benefit of utilising AI in a paid advertising strategy is that you’re hyper-targeting the most highly qualified audience: those folks who are actively seeking a solution and are ready to open their wallet for whoever can deliver.”

So, how will you deliver better PPC results to your customers in 2019? The sky is the limit – and if you don’t believe it, just remember The Jetsons! Begin the journey by deciding to make this your breakout year for machine learning adoption – to achieve better conversions, new levels of customer satisfaction, and improved ROI for your business and your clients.

Leave a Reply