This is a user generated content for MyStory, a YourStory initiative to enable its community to contribute and have their voices heard. The views and writings here reflect that of the author and not of YourStory.

AI in Search - What's behind Google's new fullistic approach?

Fullistic Setup, Automation Approach or Holistic Automation. The child has, as so often in digital marketing, several names. "Fullistic" is a word creation from "Full" and "Holistic" and means the holistic automation of campaigns in Search Marketing. But how exactly does it work?

AI in Search - What's behind Google's new fullistic approach?

Tuesday August 27, 2019,

6 min Read

Is the fullistic approach a magical lever?

Unfortunately not - Google's full-fledged approach is not a lever that you just have to kill to make all your campaigns succeed as if by magic. Rather, it is an umbrella term for the combined use of the intelligent and learning Google features that have been around for some time. 

The Google-driven character of the fullistic term is intended to express the increasingly data-driven and automated Google Ads targeting that appears in various places in the Google Ads universe. Proper use of features can automate some tasks and improve campaign outcomes using Artificial Intelligence (AI).

In concrete terms, we can identify three areas in which AI and automation are already playing a central role in search marketing: intelligent bid control, data-driven target groups, and automated ad formats. In the following, we would like to describe what is behind the individual features and how they change the search engine marketing in their interaction as a fullistic approach.

Smart bidding via Smart Bidding

Smart Bidding is Google's automated bid strategy that optimizes and controls bids at each auction through machine learning to meet given goals. According to Google, millions of signals are considered in the bidding process. For example, information about the device, the location, the browser, or the day of the week is used at the moment of a search to calculate the probability of buying a particular user. 

Especially helpful are the exclusive Google signals like the search query of a user. If the conversion probability is considered high by the learning algorithm, the bids are raised automatically, and lowered with a low probability.

Use of data-driven target groups in search marketing

Google offers us attractive target groups that we can more or less select in Google Ads at the click of a mouse. The segmentation of users is based on their online behavior. Search Marketing offers us three types of target groups that are part of the fullistic approach.

The "ready to buy target groups" are a reservoir with people who are currently very busy with a topic, do research and probably just before a purchase. This could be, for example, young men planning to buy a car. Google already offers over 500 buy-ready audiences, ranging from used compact cars to game consoles to trips to Amsterdam.

Going beyond the four demographic standards (age, gender, parent status, and household income), Google has a "detailed demographics" audience setting. There we can provide even more exact demographic conditions to our target group, for example the highest educational attainment, the age of the children of persons with parental status or the residential property status. 

On the basis of this data, a provider of real estate loans, for example, can select young families who are not yet homeowners and have sufficient income to buy a home.

With re marketing lists in Search, you can adapt campaigns for users who have already demonstrated a concrete online behavior. For example, bids can be increased or ads adjusted when a user has already visited a company's Web site and then asks for a specific search query.

Automated ad formats for dynamic content

While Smart Bidding as bid management and the target audience as the targeting instrument control the delivery of advertising, some functions are dedicated to the automated creation of advertising material.

For the creation of so-called Dynamic Search Ads (DSA), Google is currently crawling a search query and creating the appropriate headline based on a stored dummy ad. 

With a wide and often changing assortment, DSA are a great way to generate extra reach and high traffic traffic. Because ads are created automatically based on the dummies, DSAs are an important part of the fullistic approach.

The Beta Responsive Search Ads (RSA) are still in beta. Here, the Campaign Manager creates multiple headlines and descriptions in Google Ads on Google Ads. Google then tests the various combinations for their effectiveness in search queries.

For content creation, Google also offers ad suggestions. This campaign type uses texts deposited by the account manager to generate new ads. The various ads also compete against each other. Depending on the goal set by the account manager, Google selects the most successful versions.

Paradigm shift in strategic campaign building

The fullistic approach is, as I said, not a lever that revolutionizes search engine advertising from one day to the next. In order for the features and functions described above, which are based on AI and automation, to be effective, a few things have to be taken into account when creating a campaign. 

Previously, campaigns were built as granular as possible and created their own campaigns by device and target group, and created their own ad groups for match type. This was necessary to best answer every search query and efficiently distribute the usually manually managed budget.

As part of the Fullistic philosophy, campaigns should not be built too granular. It may be enough to post five keywords in the Broad Match and leave the campaign to the Google AI. If we break up the campaign into too many pieces, we are robbing the system of data base to work well. 

The campaign of the future has the following formula: simplified campaign structures, the integration of target group signals, smart bidding in combination with data-driven or position-based attribution and the correct ad setup including responsive search ads as well as additional dynamic search ads. 

The result: higher click rates and quality factors, lower click costs. The bottom line is that in the perfect scenario with the quality of the campaigns, sales will increase with improved efficiency.

Data is the fuel with which everything stands or falls

Whether it's smart bidding, using data-driven audiences or dynamic ads, we need to feed the system with data to make it work well. This can be done in a variety of ways, from tracking visitors to our website to uploading CRM data to adding offline conversions as part of omnichannel campaigns. Basically, the more data, the better. For example, when setting up Smart Bidding, Google indicates that the automated bid strategy will not work properly unless there are at least 30 conversions (50 target ROAS conversions) in a one-month period.

What changes for advertisers and agencies?

Of course, automating some tasks saves the campaign manager time. The account manager is increasingly becoming a consultant, helping advertisers identify the relevant audiences, appropriately combining the smart features and constantly analyzing the results.  Agencies are increasingly in demand to provide advice to their customers across channels. 

After all, Google Ads alone combines the disciplines of search engine advertising (SEA), display advertising (GDN) and social advertising & influencer marketing (YouTube). All these measures need to be coordinated. In addition, the Google forge is constantly producing new features that need to be evaluated and tested by experts in agencies. The consulting service,

All in all, AI makes search engine advertising better, but human campaign managers are by no means superfluous. The strategic advice on the use of technologies and the operational experience in the application make agencies in the future an important partner for advertisers.