How Smart is it Really?
Google Conversion Optimizer, automated bidding, smart bidding. Google Adsâ€™ free-to-use, in-house bidding tools have been labeled different names over the years, with one overarching design concept: to alleviate PPC managers from the complexity of bid management using tools built-in to Google Ads.
Google offers many bid strategies and has continued to expand that list over time. Current hallmark strategies include Maximize Conversions, Target CPA, and Target ROAS, among others. Requirements? The more important, conversion-based strategies require accurate conversion tracking using the Google pixel, but outside of that Google strategies are available to any program.
Why would Google want advertisers to use their own bid optimization options? Is it to offer an out-of-the-box convenient solution to inexperienced advertisers? Or is it to straddle the line between improving conversion rates and increasing CPCs? The speculative answer is both. Rumor has it that Google poured development resources into their bidding capabilities (once viewed as a poor strategy) to accommodate the many thousands of small-time advertisers who may not know much about digital ads and who have more important things to focus on. Mission accomplished; so well, in fact, they got the attention of large advertisers with dedicated PPC teams and vendor relationships. Where do things stand now?
Auction time bidding
This feature has gained recent publicity, but itâ€™s been the case for a long time. Google rightfully recognizes it as the premier benefit of their own bidding solutions. As the publisher serving the ads, Google can interact with search queries, including bidding on each unique search query, rather than having a single bid per keyword like all other bidding options are tied to. This means the relevance of every search query is evaluated and bid on in real-time. For broad match keywords, this capability can be especially valuable, since the relevance of search queries can vary widely.
|Search Query||Keyword||Auction Time
|â€˜household maid serviceâ€™||â€˜household cleaning serviceâ€™||$3.50||$1.50|
|â€˜best household cleaning suppliesâ€™||â€˜household cleaning serviceâ€™||$1.10||$1.50|
Google claims to use the full suite of user data. There are still many limitations to Googleâ€™s array of user data, but it grows every day. The majority of that data is available to utilize by anyone via bid adjustment modifiers. However, much like with auction time bidding, there are some data points that only the publisher serving the ad can interact with at the time of auction. A couple of examples are browser and operating system. Thereâ€™s no consensus about how impactful this data may be, but since only Google can make use of this data in real-time, it is another bidding advantage.
Simpler Campaign Management
This is also related to auction time bidding. Because Google can bid on each search query, thereâ€™s no need for the use of different match types. A campaign with only broad keywords would theoretically perform as well as if that campaign also had phrase and exact match. This may be especially valuable to smaller advertisers that lack budget or resources to build and maintain a larger, more sophisticated, ad account.
Lack of Transparency
With Google uniquely bidding on every search query, an advertiser has no access to the complete bid history of any keyword or any individual auction. Keyword level CPCs are not reflective of auction dynamics, given that the bids for each search query may be incredibly wide-ranging, even for the same keyword. So, the CPC for a given keyword acts as the average of all the unique bid auctions, and that is the most information available. For advertisers looking to analyze granular performance, the possibilities with Google are limited.
Lack of Control
Using Google bidding means relinquishing control across different market segments to what Google interprets as the most likely to convert. Thatâ€™s fine in a vacuum, but an advertiser will commonly know more about their business than whatâ€™s reflected in historical online conversion metrics. For instance, a business is set to release a new mobile landing page or has a call center with variable staffing availability. Google can only react to how those efforts change conversion rates over time, which is too late, and businesses miss out on influencing website traffic based on important information that Google cannot incorporate.
Data Optimization Limitations
Google bidding is restricted to relying on conversion data tracked by the Google pixel. This means a user is unable to incorporate extraneous or contextual data into bid optimization. Manual or vendor bid strategies can layer in BI data such as product inventory, or contextual data such as weather. For many businesses, performance is closely tied to external influences, and itâ€™s best to take advantage of that data wherever applicable.
Insufficient Data Volume
Included in the Google Recommendations tab is a recommended list of campaigns to apply smart bidding (typically with Target CPA). This list can be downloaded and includes an estimated performance impact for each campaign. The catch is that not every campaign will be recommended for smart bidding, usually because Google requires a minimum threshold of conversion volume to achieve confidence in performance. Google projects when itâ€™s unlikely to be effective and itâ€™s not designed to accommodate campaigns with limited volume. Depending on campaign size and conversion rates, this gap in coverage can make up a significant portion of spend.
We at QuanticMind have seen time and time again a pattern that Google bidding increases CPCs. Belief is that Google is aggressively targeting and bidding up relevant searches, but if that increase in CPC is not met with an increase in conversion rates, CPAs go up and margin goes down. This inflation is also everlasting. Hypothetically, if more and more players in the auction rely on Google bidding, the net impact on auction dynamics results in a higher cost to generate the same amount of clicks, and everyone simply pays more to Google.
When to use Smart Bidding?
Itâ€™s best to understand the trade-offs using smart bidding versus a manual strategy or a bid management platform like QuanticMind. Know what is being given up by relying on Google bidding using only Google tracking. Thereâ€™s an obvious theme to relying on smart bidding – you bought a car from Google and you are having them drive you around in it. If youâ€™re an inexperienced driver or arenâ€™t sure where you want to go, this is an effective solution, but for anyone who wants the ability to input direction, youâ€™re ceding complete control.
A Hybrid Approach
The seasoned advertiser will likely understand what bid strategy to utilize for different segments of their program and business. With a proper understanding of the Pros and Cons of Google automated bidding, an advertiser can test and recognize the ideal strategies. To achieve peak performance, that may include the use of myriad strategies. And it makes sense, X + Y is better than X or Y in isolation, once you understand where X > Y and Y >X. Rarely is a bidding solution one size fits all. Keep on learning. Google may or may not be the right solution.