Automation in performance media campaigns offers efficiency and scale, but relying on it blindly can lead to wasted spend and missed opportunities. In this deep dive, Kevin Sung, our Digital Performance Director, uncovers the hidden pitfalls of unchecked automation and shares strategies to ensure your campaigns stay on track.
Almost fully automated formats like Google’s Performance Max (PMax) and Meta’s ASC+ campaigns have emerged as powerful allies that have, in recent years, been adopted over other equivalent or overlapping formats. These formats promise efficiency and scalability through machine learning. However, entrusting your campaigns entirely to automation can lead to unforeseen pitfalls, not always evident when reviewing the data in reports native to the platforms.
Let’s delve into the potential hazards of unchecked automation and explore strategies to navigate them effectively.
1. The Peril of Promoting Underperforming Products
Automated bidding strategies, such as PMax, are engineered to maximise conversions. Yet, without vigilant oversight, they may channel budgets toward products that attract clicks but fail to convert, resulting in wasted expenditure.
Search Engine Journal highlights some key considerations when setting up PMax campaigns. While PMax can drive conversions, it may also lead to performance volatility if not closely monitored, resulting in inefficient budget allocation. The key takeaway from this point is to consider that even if your campaigns are achieving the target ROAS or conversions, you should still review product and category-level investment. This process can uncover pockets of significant wasted spend.
This behaviour underscores the importance of looking beyond surface-level data. For instance, a product that garners significant traffic but doesn’t contribute to your overall profit margin should be flagged and investigated. Addressing such inefficiencies might involve adjusting your product feeds, revising your bid strategies, or implementing performance filters to exclude underperformers.
Additionally, using advanced attribution models can help determine whether these products play a role in assisting conversions elsewhere in the funnel. If no such role is evident, this further solidifies the case for exclusion.
Regularly scrutinise your product performance data. Identify items with high spend but low conversion rates and adjust your campaigns accordingly. Implementing negative keywords and excluding underperforming products can help reallocate the budget to more profitable items.
It has long been known that PMAX can deliver questionable placements, formats, and surfaces, as Search Engine Land and performance marketing world highlights, potentially leading to inefficient budget allocation.
Specifically, PMax campaigns often distribute budgets across various channels, including display and video placements. While these channels are valuable for brand awareness and salience, they may not always yield the highest return on ad spend (ROAS) compared to search ads or PLA.
Using reporting dashboards or custom scripts to break down performance by channel and format, advertisers can identify trends indicative of over-delivery. If these placements aren’t driving sufficient conversions, a strategic realignment is necessary. This could involve creating exclusions or reallocating budgets to formats with higher efficiency.
Implement close-to-real-time monitoring using customised reports and scripts to track where your ads are being displayed. If you notice over-delivery in display or video placements without corresponding conversions, consider course-correcting actions such as limiting assets, reducing investment, or shifting targets.
Automation tools operate based on the data they receive. Without integrating comprehensive business data—such as stock levels, profit margins, and seasonal trends—these tools might promote products that are overstocked, out of stock, or less profitable.
Econsultancy emphasises the importance of taking control of your Google Shopping feed. Many advertisers treat it as a ‘set and forget’ channel, leading to inefficiencies. However, the point goes beyond this and speaks to understanding, modelling, and predicting sales trends.
The integration of business data is often underutilised in automated campaigns. Advertisers should consider dynamic data feeds that include variables like stock availability, profit margins, and dynamic pricing. For instance, if certain products have a higher profit margin during specific seasons, this data should directly influence bidding decisions.
Furthermore, predictive analytics can be incorporated into your strategy. This allows advertisers to proactively adjust campaigns in anticipation of trends rather than reacting to real-time data alone. By aligning advertising activity with anticipated demand, marketers can achieve a balance between efficiency and responsiveness.
In summary, Integrate your business data with your advertising platforms. By providing detailed product information and aligning your campaigns with inventory levels and profit margins, you can ensure that automation tools make more informed decisions, leading to better performance.
Automated campaign formats often apply uniform strategies across diverse products and audiences, potentially leading to inefficient spending.
Purely building asset groups around audience signals in PMax campaigns can confuse the system, implying that a one-size-fits-all approach may not be effective. However, implementing segmentation can be challenging when dealing with limited conversion data. To overcome this, advertisers may need to aggregate data across multiple time periods or experiment with different segmentation variables.
So, where conversion data permits, segment your activity by category, price bracket, or any other horizontal segmentation that aligns with your objectives. Create tailored strategies for different product categories, audience segments, or geographic locations.
This approach allows for more precise targeting and budget allocation, enhancing overall campaign performance. As mentioned, the caveat here is data volumes. Specifically, ensure each split has enough conversion data to power the machine learning algorithms.
While AI and automation excel at processing large datasets and optimising for specific metrics, they may lack a nuanced understanding of market dynamics, competitor actions, and emerging trends.
It often lacks the ability to react to real-time market dynamics outside its trained model view. Using competitor insight tools like Adthena can provide context for the broader market outside of the activity you appear on. Adthena refers to this as their ‘whole market view’.
Build in a process of regularly reviewing market trends, competitor strategies, and consumer behaviour. Use these insights to inform your automated campaigns, ensuring they align with the broader market context and capitalise on emerging opportunities.
While automated and AI-led solutions offer significant advantages in managing complex advertising campaigns, blind reliance on them can lead to inefficiencies and missed opportunities. By integrating human insight, regularly analysing performance data, and customising strategies to align with specific business goals, advertisers can harness the power of automation while mitigating its potential drawbacks.
The most successful campaigns will be those that blend the efficiency of automation with the strategic acumen of human oversight and data that the systems haven’t had sight of or been trained on.
Here at Bonded, we’ve mastered the blend of automation & human input and developed an auditing solution that allows us to effectively analyse activity in detail. Interested in understanding if you are maximising your budget? Reach out to us today to arrange a complimentary PPC audit and consultancy call.
Performance Max for Ecommerce: Advanced Strategies and Pitfalls to Avoid, Search Engine Journal
Google Performance Max: Everything you need to know, Search Engine Land
Peeking behind the curtain of Google’s Performance Max, Performance Marketing World
PPC: When automation may not be the best option for retailers, Econsultancy
