Automated vs. Manual Bidding: My Professional Insights from my Adventures (and misadventures) 

Automated vs. Manual Bidding: My Professional Insights from my Adventures (and misadventures) 

Automated vs. Manual bidding - LeapOut

Strolling down memory lane, I think back to the good old days of digital marketing. Days filled with Excel spreadsheets, data analysis, and meticulous manual targeting. From sunrise to sunset, we crunched massive amounts of data, determining the most appropriate CPC (Cost Per Click) bids, and unraveled the mysterious science of driving conversions. The hustle was real, but the gratification of boosting sales at a lower CPA (Cost Per Acquisition) was unparalleled. The only taste of automation we had back then was the simple, yet efficient Bulk upload editor. 

With the advent of machine learning and artificial intelligence, the digital marketing landscape is transforming at a staggering pace. The hands-on, granular control we used to have been gradually shifting towards automation and AI (Artificial Intelligence). And as we stand on this precipice of change, it’s essential to assess when to embrace automated bidding and when to stick with the old faithful – manual bidding.

My experience has shown that automated bidding can be your best ally when it’s used persistently and prudently. It’s not a set-it-and-forget-it tool, but rather a dynamic aide that requires fine-tuning and constant monitoring. When we tested Google’s Smart Bidding on our client accounts, we witnessed a significant improvement in performance. It felt like unearthing a hidden treasure in the quest for conversions. AI was our knight in shining armor, consistently delivering enhanced campaign results. 

However, there’s a caveat. Just recently, I ran a Click to Messenger Ad on Facebook with the objective of attracting business inquiries. The outcome was baffling. My inbox was flooded with messages from unqualified leads, majorly the elderly from remote provinces in the Philippines. I figured that the audiences had a high propensity to click on messages (not to talk to an agency) but were not potential clients for a digital marketing agency. That experience was a stark reminder that machine learning, while efficient, is not always effective, especially if you lack data.

So, when is automated bidding should be used on Google ads and Facebook ads?

  • Ample Conversion Data: Automated bidding works best when it has a large pool of conversion data to learn from. This allows the algorithms to better understand the trends and patterns associated with successful conversions. If your Google Ads or Facebook Ads campaigns have been regularly accumulating significant conversions (recommended at least 30-50 conversions in the past 30 days for Google Ads and at least 50 conversions per week for Facebook’s Conversion Optimization), then automated bidding can be highly effective. 
  • Time Constraints: When you are pressed for time and can’t afford to constantly monitor and adjust your bids, automated bidding comes to the rescue. It shoulders the responsibility of real-time bid management, freeing you to focus on strategic tasks like developing compelling ad copy, refining your targeting, and strategizing your campaign. 
  • Driving Conversions: If your main aim is to ramp up conversions or maximize conversion value within a specified budget, automated bidding strategies can serve you well. Facebook’s “Cost Cap” and “Bid Cap” strategies and Google’s “Maximize Conversions” or “Target ROAS (return on ad spend)” strategies leverage extensive data and machine learning to predict which ad placements are likely to result in conversions or high conversion value. 
  • Venturing Into New Territories: Whether you’re experimenting with new keywords, breaking into new markets, or targeting new audiences, automated bidding can fast-track your data collection. It allows you to experiment with diverse bid amounts efficiently, thereby expediting the learning phase of your campaigns. 
  • Managing Complex Campaigns: If your campaign management involves multiple keywords, a myriad of ad groups, or complex targeting configurations, automated bidding can significantly simplify your task. Machine learning algorithms can adjust bids at the granular level of individual keywords or targeting settings—a feat that would be overwhelmingly time-consuming if done manually. 

When should you do manual bidding?

There’s no one-size-fits-all answer to this question. However, instances when automated bidding falls short provide a compelling argument for manual bidding. Take my recent tryst with Google Ads, for instance. We were receiving many unqualified leads, such as job applicants and suppliers. While these leads might have used the right keywords, they weren’t the conversions we were aiming for. But the algorithm was registering them as such and chasing these clicks, causing our budget to dwindle. In such scenarios, manual bidding offers a level of control that automated systems sometimes lack.

For me, marketers should opt-in for manual bidding during these scenarios:

  • Always-on Campaigns: Have you identified those money keywords or specific audiences that consistently drive profitable engagement to your campaigns? These are your money keywords or audiences – the ‘always-on’ and brand keywords that you need to maintain a firm grip on. In these cases, manual bidding gives you the reins to control ad positions effectively. This approach allows you to assert your dominance, ensuring that your brand remains at the forefront of relevant search results. 
  • Data Acquisition Phases: Machine learning, for all its brilliance, is only as effective as the data it must work with. If your data pool is lacking in quality, your conversions will reflect this inadequacy. As the saying goes, “garbage in, garbage out.” Here, manual bidding comes to the rescue. It allows you to actively steer your campaigns towards acquiring the right kind of data, giving machine learning the correct base to enhance both the efficiency and effectiveness of your campaigns. 
  • Data Reset: There will be moments when you realize that your automated bidding strategy hasn’t been working as expected. It has been chasing the wrong leads or inflating your costs. Contrary to widespread belief, this is not the end of the world. These moments of error provide opportunities for learning and growth. Switching to manual bidding in such instances lets you reset the data, offering a fresh start. You can recalibrate your campaigns, hone your targeting, and reestablish the path to achieving your desired outcomes. 
making automated and manual bidding

Making Manual and Automated Bidding Strategy Work Together

Choosing between manual and automated bidding can sometimes feel like an intricate balancing act. Each has its merits and drawbacks, and understanding when and how to use them is key. However, it’s not always a black-and-white choice. Combining manual and automated bidding can often yield the best results, provided they’re used effectively. Let’s explore the strategies, steps, and tricks to harmonize these two bidding approaches and maximize your campaign performance. 

1. Understand the Nature of Your Campaigns

It’s crucial to comprehend that not all campaigns are created equal. Some campaigns are more stable, with predictable performance, while others may be more volatile due to various factors such as seasonal trends or competitive landscapes. For stable campaigns with a wealth of conversion data, automated bidding could be a great option. On the other hand, volatile campaigns or those with sparse conversion data might benefit more from a manual bidding strategy. I’m a control freak so that’s which points to my second tip.  

2. Start with Manual Bidding

When launching a new campaign, consider starting with manual bidding. This allows you to gain a better understanding of the campaign’s performance, how certain keywords perform, and what kind of CPCs you’re dealing with. This information can be invaluable when you switch to automated bidding, as it will provide a benchmark and let you better evaluate the performance of automated bidding. I usually test ad groups and audiences differently and whenever I feel that I have a good data representation and it can be scalable, then you can start to trust that automated bidding can deliver.  

3. Gradually Transition to Automated Bidding

Once your campaign has a substantial amount of conversion data, and you’ve identified trends and insights from your manual bidding strategy, you can start transitioning to automated bidding. Similar to Facebook, Google’s automated systems work best when they have a large amount of conversion data to learn from.  

4. Regularly Monitor and Adjust

Whether you’re using manual, automated, or a combination of both bidding strategies, regular monitoring is essential. Automated strategies may require less day-to-day management, but they still require oversight. Keep an eye on your performance metrics and be ready to step in and make manual adjustments when necessary. This could mean adjusting your target CPA or ROAS, or even switching back to manual bidding if the conditions require it. 

5. Use a Hybrid Approach

For Facebook Ads, the process of ad creation and optimization requires both a systematic and creative approach. Before we fully entrust our campaign to Facebook’s Lookalike Audience (LLA) feature, powered by machine learning, we need to ensure the data is as clean and as high-quality as possible. Herein lies the importance of what I like to call the “rinse and cleanse” method. 

The “rinse and cleanse” method is like a purification process for your audience data. We start by creating ad sets for each audience expansion. Each ad set serves as a distinct playing field for a particular audience group. This approach provides a clear picture of each audience’s performance, enabling us to assess and adjust our strategies effectively. 

As the campaign progresses, we begin finding outliers – those users within our audience sets that don’t align with our ideal customer profile or those whose interactions are not contributing to our campaign objectives. The “rinse” part involves systematically removing these outliers from our audience sets. Think of it as pruning a tree – cutting away the parts that don’t contribute to its growth to allow for healthier development. 

Once we’ve done a thorough “rinse,” we move to the “cleanse” phase. This stage involves creating a data association of the performing audiences. We look at the characteristics of the users that are positively impacting on our campaign performance. By recognizing these trends, we can further refine our audience sets, focusing on those user groups that are most likely to contribute to our campaign objectives. 

Only after we’ve gone through this “rinse and cleanse” process do we let machine learning take the reins. By this point, our LLA feature has been provided with high-quality, streamlined audience data. This ensures that the machine learning algorithms work in our favor, using reliable data to find similar high-performing users within Facebook’s vast user network. 

6. Don’t Forget About Manual Overrides

While it might seem counterintuitive in an automated setup, manual overrides can be highly beneficial. For instance, if you notice that certain keywords or audiences are underperforming or costing too much per engagement, you can make manual bid adjustments to these keywords. This allows you to retain some control while still enjoying the benefits of automation. This is effective specially if you have clients who don’t care about data-driven bids but would want to always show their ads up as I mentioned in the scenarios above. 

7. Refine and Retest

Finally, always keep refining your bidding strategies. What worked once might not work indefinitely. Regularly retesting and tweaking your approaches is the key to staying ahead. You may find that different strategies work better at different times, so be open to switching between manual and automated bidding as needed. There will certainly be a point when ad fatigue or audience saturation happens. When this time comes, it’s not the end of the world and certainly you can start again.  


In Conclusion

To wrap it up, AI and machine learning are powerful tools. They adapt, learn, and can make your campaigns more effective. But like all tools, they’re not perfect and sometimes may not make the best decisions for your specific business needs. When this happens, don’t hesitate to step in and adjust. 

Digital marketing is a thrilling journey, full of constant learning and adapting. While AI and machine learning represent the future, don’t forget about manual bidding. There’s a time and place for both, and it’s important to stay flexible. Keep learning from your successes and mistakes, and adjust your strategies as needed. 

Remember the phrase, “Garbage in, garbage out”? This rings especially true in automated bidding. If you’re going to use automated bidding, make sure your data is accurate. Otherwise, you might spend a lot of time cleaning up and filtering data. Regardless of the advancements in technology, the results you get are only as good as the data you put in. 

But you don’t have to go at it alone. At LeapOut, we specialize in both Google Ads and Facebook Ads. Our team can help you make the most out of both manual and automated bidding. So, don’t hesitate to reach out if you need help. We’re ready and eager to assist in boosting your digital marketing campaigns. 

Marvin Ortiz
Marvin Ortiz

Marvin Ortiz is a digital entrepreneur with a wealth of experience working with local and international brands as an e-commerce and digital growth expert. He is the Founder and Co-Owner of LeapOut, the Philippines leading E-commerce SEO and Digital Marketing Agency. He loves to read and hit the beach during his free time. You can get in touch with him on Twitter at @marvinortizph.