Hero image

Ways AI Could Improve PPC Performance in 2023

Written by Joe
Get in touch

Are you using AI (Artificial Intelligence) in your PPC marketing strategy? 

With the new ChatGPT (which we are sure you will have heard of) taking the world by storm this month, maybe it’s time to dive into the world of AI within your digital marketing plans. 

There’s an abundance of automation we already use within PPC and have been using for quite some time in terms of making the paid process simpler, more efficient and more effective. Whether these come under ‘AI’ or more of an automation process is an argument to be made, however in this blog, we share how and where these processes can be used within Paid Search. 

ChatGPT Takes The World By Storm In 2023

ChatGPT is something we (and everyone else) have already spoken about. If you haven’t checked out our recent blog on it, you should – ChatGPT – what’s all the fuss?

The ChatGPT tool provides some great opportunities for ideas and creativity when you may be hitting a mental block, or you may just need a little bit of inspiration. 

This is an example we put together on mattresses for a paid ad: 

We all spend a lot of time writing ad copy and due to its nature, it has to be encapsulating and something that draws the customer in when the SERP is full of other potential competitors. The obvious positives here are what is mentioned above, help with inspiration, and help writing huge lines of ad copy but like many things in AI, there will always have to be some form of human input to ensure that things don’t slip the net.

Autonomous Processes We Use Within Paid

Google is forever pushing the ‘new and improved’ smart bidding, a machine learning system that devises bid strategies based on the data available to them within the Google ads accounts. 

The 4 main ones here are: 

  • Maximise Conversions
  • Target CPA (Cost per acquisition)
  • Target RoAS (Return on AD Spend)
  • Enhanced CPC (Cost per click)

Within the PPC team at Flaunt, we mainly use a wide range of the above automated strategies, however, the target RoAS strategy is prominent across an array of our ecommerce clients. We tell the system what RoAS we would like to see from that campaign (as an example we want to hit a 500% RoAS), the system then uses data available from the account to push products within that campaign that it believes is going to help us achieve that figure.

For example, if a client wants to push items that are of a higher value in their shopping campaigns. As these campaigns are on Target RoAS strategies it will automatically push items that sell well, in this case, they may be cheaper items because they sell a lot of volumes, so whilst the individual item prices are lower the volume is significantly higher. We opened the Target RoAS strategy up (from 500% to 300% as an example) to allow the system to spend a bit more, gather a bit more data and ultimately push those higher-priced items. 

There are also a number of automations that we can look at from a paid perspective that help us analyse and improve efficiency and overall performance. An example would be heatmap scripts, set up to automate analysis of how customers are behaving through our paid activity. This gives us data on what time of days our campaigns are most efficient, what days of the week, what locations etc. We use this data to make manual adjustments but, there are ways to automate that process to then adjust bids based on what is seen across those scripts. 

The human element is still hugely important within these automation processes as we need to know what to feed the AI to leverage the best results and how to segment things to still allow the human to pull levers to impact performance. 

Automated Bid Adjustment Scripts

Within PPC we use scripts as a safety net for when things may not go our way! They give us the ability to pause poor-performing ads and keywords, by simply setting up automated rules that will automatically pause ads based upon the rules we set. For example, we can set a rule to pause any ads within a campaign that in the last seven days has not received more than one conversation. 

With anything like this, it is significantly top-level –  there may be a reason that the campaign hasn’t converted, maybe due to an issue we haven’t picked up on yet and by pausing it, we’re not fixing the problem. However, it could be that the ad itself is performing poorly and pausing it is the best solution. 

Automated rules also let us know when we have any downtime on ads. The last thing we want is for paid ads to by pushing people to a site that is currently not working or having significant downtime. That’s why we use automation to set up alerts to let the paid team know if certain sites have gone down and as a result needs to be paused. 

A Last Word From Us On AI and Automation Within PPC 

The main talking points here are simply put, AI and automation are great and we use them day in and day out to help us with efficiency and processing. However, there are fairly obvious restrictions that also mean we need to still have a lot of human input. 

With smaller accounts or if you were managing your own business activity then the automated bids and strategies are something that would really appeal to you. For significantly larger accounts automation is just a small cog in a very large team effort to ensure we are seeing the best performance possible.

Want to know more about AI and digital marketing trends? Let us know on LinkedIn, or, if you’d like to check out some other industry insights, keep an eye on our blog and subscribe to our Flaunt newsletter.

more like this.