Google Implements Deep Machine Learning, New Mapping Tech for Measuring Store Visits

New store visits measuring innovations will give new advertisers these insights, and a richer more complete data set to advertisers already using store visits.

Share online:

A series of new innovations that will make store visits available to more advertisers around the world, as well as improvements in how store visits are measured and reported in AdWords, will give advertisers already using these insights more complete view of performance was announced on Wednesday.

Since introducing store visit measurement two years ago, Google AdWords has measured over 4 billion store visits around the world. Here's how it will improve store visit measurement going forward.

More Richer Data at Same Accuracy

As measurement technology evolves in machine learning, mapping technology and survey quality, Google says, AdWords will now increase reporting more number of store visits, while still achieving same high bar for accuracy.

Deep Learning

As deep learning is now used to measure store visits, this change will allow Google to measure more store visits in challenging areas, such as in multi-story malls and dense geographies where many businesses are close together.

New Mapping Initiatives

Google says they're now able to map the precise geography and geometry of business locations. This was accomplished through Google Earth and Street View's refreshed imagery, and scanning Wi-Fi signal strenght inside buildings to figure out their exact doundaries.

Survey Data for Store Visits Verification

To verify when a potential visits occurred, the Google systems conducts a survey of select users about their store visits and ask about the locations they've visited. This data is then use to calibrate machine learning models. This strengthens the accuracy and allows Google to report more store visits that otherwise may have been excluded previously.

"We've continued increasing the reliability and quality of these surveys using real-life audits and location visits from our own teams," the team said, for example, "we're now verifying more visits that happen in areas with high store densities to give our models more data to learn from," the team said.

In other news, Google App Engine flexible environment is now available in the europe-west region.

To get started, simply open up the Developers Console and create a new project, and select App Engine. After choosing a language, you can now specify the location as europe-west. Note that once a project is created, its region cannot be changed.

Google App Engine Flexible Enviroment Availability Map
Google App Engine Flexible Enviroment Availability Map

Google with Komprise created a solution guide "Archive your cold data to Google Cloud Storage," that describes how customers can understand data usage and growth in their storage environment, get a customized ROI analysis and move this data to Cloud Storage based on specific policies.

See the full tutorial guide here.