To get the best possible results with Google Ads, marketers create sophisticated campaigns based on search terms, products, users' location or devices. Now even more efficient results can be achieved with the new AI-driven Google algorithm, that delivers data-driven decisions and better campaign results with the same budget. Best of all, these results can be topped with the Triple-A Approach (Audience, Automation, and Attribution) that has been incorporated into the algorithm, developed by the pure Google Ads agency Smarketer.
High-quality AdWords campaigns are characterized by their granular and thus highly segmented structure: Individual campaigns are designed manually, depending on the product or brand keywords, the gender and location of the users, and the various end devices. Smarketer’s Triple-A Approach serves the new Google algorithm and delivers innovative levels of power through machine learning, which the AdWords agency explains as follows:
Audience: away from the search terms - to the target groups
Instead of search terms and keywords, the new approach focuses on the users themselves, who are looking for shops or products. Hence, the focus shifts from the keywords to the target groups. Thanks to machine learning, the algorithm can analyze up to 70 million signals within 100 milliseconds and evaluate all possible data combinations in the shortest possible time. This ongoing processing and re-prioritization of audience information analyzes which users are most likely to make a purchase in the future. Google updates this information daily by adding or removing users in the audiences, but information that engages users in a context is hugely important to advertisers.
Automation: break up the granular structure - and let the algorithm work
Automation means primarily bid optimization, the so-called Smart Bidding, for which dealers can also use optimization tools from Google. The benefit of the new Google algorithm is that merchants can dramatically improve their performance, since Google processes different signals in real-time when serving an ad now – including search queries, browsers, age, gender, interests, devices, location and date, in other words information that is not included in a manual control. So far, dealers had to set their bids manually. Nowadays the Google algorithm can control data-bases, without subjective opinions, not to mention that Machine Learning achieves much better results than the previous granular structure.
Attribution: not only the last click counts - but the path getting there
Until now, dealers mainly focused on the last click when it came to the evaluation of the conversion - the one that leads to the sale. However, the attribution model should be changed from "last click" to "data-driven" or "position-based". For example, if the customer journey led from Facebook to Google and from there to the dealer's landing page, according to the old model, sales were made exclusively by Google. If dealers change their attribution model to "data-driven," the value of the conversion is distributed to the action. Thanks to artificial intelligence, the optimal model for each customer can be calculated, which assigns the different campaigns and search terms. It shows the road to the purchase, since users generally need multiple attempts before converting. In addition, if there is not enough data available for the data-driven attribution model, the position-based model can be used, which allocates 40% of the first and last viewed ad, and the remaining 20% to the remaining clicks.
Clearly, the AI-driven Google algorithm makes it necessary to merge and downsize campaigns, because over-segmentation blocks machine learning.
By the way, this best practice from Smarketer, Google's multi-award-winning cooperation partner, has been included in Google's Agency of the Future (AOTF) program. Obviously, Smarketer understands well how to build campaigns to use the algorithm efficiently and use smart bidding properly.
By Daniela La Marca