Navigating Google Ads (PPC) in the AI era – The death of manual lever-pulling
Why AI Google Ads Management Is Shifting To Governance
The goal of machine learning and AI Google Ads management is to make managing this complicated marketing channel simpler and more efficient, leaving you free to focus on the wider strategy instead of digging around in accounts all day.
It all started with manual bidding at the keyword level, but PPC hasn’t really been a fully manual marketing channel for quite a while now. It has utilised machine learning for bid strategies such as Enhanced CPC (eCPC) for many years. Smart Bidding then came into existence, with levers like tCPA and tROAS targeting becoming popular optimisation methods.
Now, the wider online landscape has evolved as well, with AI becoming the talking point across many aspects of our lives. We have LLMs like ChatGPT, Gemini, Claude and others being used to share knowledge, provide instruction, give feedback and much more. Then we have the evolution of the search engines themselves, with features like AI Overviews and AI Mode in Google search popping up.
With these changes, running PPC campaigns in the current landscape means utilising features like broad match keywords, Performance Max (PMAX) campaigns and/or AI Max if you want visibility in these AI spaces. At the time of writing, it is not yet definitive which of these individual features are necessary to show up in AI spaces, or whether they all need to work together as a set. However, it is worth testing each area to determine which is the best fit for your business.
Guarding The Black Box
Performance Max campaigns are nothing new. This campaign type has been running in Google Ads for several years now. What started as something of a black box campaign type, has evolved over time, with Google introducing key features that make it more transparent. The Channel Performance report provides visibility into which sub-channel (e.g. Search, YouTube, Gmail, Maps, Display, Discover, Search Partners, Shopping) receives the highest share of spend, allowing you to identify where optimisation efforts should be focused.
For example, if the YouTube sub-channel is receiving a large share of impressions and spend within your PMAX campaign, consider optimising your video assets. If Search is driving most of the activity instead, think about refreshing your headlines, descriptions and other creative assets.
Pro tip: You need to understand if your PMAX activity is cannibalising your Search campaigns. If it is, you need to compare the performance of each campaign type and determine which should be your primary focus.
Running paid ads today may seem increasingly daunting as the platforms continue to evolve, but if essentials like conversion tracking and campaign structure are implemented correctly, you’ll be in for a much smoother ride.
When it comes to conversions, make sure you’re tracking the right conversion action for your business, whether you’re a service business generating leads or an e-commerce store pushing for purchases. Whatever your primary conversion may be, you need to give the Google AI both the correct signal and a sufficient volume of that data in order to optimise performance accurately.
On the surface level, smart bidding may seem quite simple for some (you give Google a target, it aims to hit it) but for others it may be a little black boxy too – To utilise Smart Bidding strategies like tROAS, it is generally recommended that campaigns generate at least 30 conversions over a 30-day period. For tCPA, the recommendation is at least 15 conversions over the same period. These are recommended minimums because Google’s algorithm requires enough conversion data to effectively learn how to bid towards your target ROAS or CPA. Additionally, with the higher volume of data, you yourself can better understand how other metrics are performing when Google tries to reach the set target.
That said, we can’t simply rely on Google to do all the work. Google Ads isn’t a platform you can simply set and forget. You still need to consistently add negative keywords, rotate out poor-performing assets, A/B test ad copy, monitor keyword performance and much, much more. AI can help with a lot of this but you still need to be there to oversee).
Advanced Strategies: The Creative Lever
Optimising Paid Search campaigns in the AI era has shifted more towards keywordless technology and campaign types. Historically, advertisers prioritised exact match keywords because of their efficiency, often pairing them with phrase match to capture additional volume. Broad match was traditionally the match type many advertisers avoided, preferring modified broad match instead.
However, as Google Ads and Microsoft Ads have evolved, their machine learning and AI capabilities have improved to the point where broad match keywords are now considered a cornerstone of many PPC campaigns. Broad match helps capture significant search volume (without marketers having to manually find all keyword possibilities themselves to bid on), while Smart Bidding controls the efficiency to help campaigns remain on target. As a result, phrase match has become less commonly used.
Now, we’re moving further towards a keywordless era of PPC. For example, campaign types like Performance Max (PMAX) don’t rely on keywords in the traditional sense. Instead, they use AI and signals gathered from creative assets you provide such as landing pages, product feeds and asset groups to understand where ads should be shown.
To ensure your campaigns appear in the placements that matter most to you and your business you need to regularly optimise the creative layer alongside optimising your traditional search layer. So when optimising keywordless campaign types like PMAX, don’t just focus on headlines and descriptions. Ensure you’re also optimising visual assets such as images and videos, as these play a significant role in helping Google’s AI understand your offering.
Technical Grounding In Data
Maintaining a strong and consistent flow of conversion data is essential in today’s paid advertising landscape with AI at its core. Without it, platforms like Google Ads and Microsoft Ads may optimise towards the wrong signals. This can result in campaigns targeting users who cost your business money but deliver little to no value, whether that’s in revenue or meaningful data.
Give ad platforms the right information and they’ll become much better at finding more of the users who provide the greatest value to your business. (The AI can do a lot for improving account performance but only if you provide it with accurate information).
Final Thoughts On AI Google Ads Management
Gone are the days when advertisers had complete manual control over every granular aspect of their paid advertising. The future of PPC increasingly relies on machine learning and AI, both of which already play a significant role in day-to-day campaign management and are likely to become even more influential over time.
To avoid falling behind, you need to adapt to the ever-changing PPC landscape and embrace hybrid marketing workflows. While this may seem daunting, it’s important to remember that AI is there to assist you. It’s a tool that follows your direction, not one that takes complete control.
To steer it effectively, remember a few key principles:
- Give it accurate data and a consistent flow of it.
- Provide accurate conversion signals so it can learn what your business values most.
- Set realistic targets when using tCPA or tROAS.
- Stay on top of Search Terms reports and regularly add irrelevant queries as negative keywords.
If you keep these principles in mind, you’ll allow AI to handle much of the day-to-day campaign management while you focus on providing the strategic direction.
Top-Level Comparison – The Old vs. New:
