How digital advertisers can navigate the updated cloud solutions from Google

Google recently announced BigQuery Omni, a solution bringing multi-cloud analytics to data. This solution allows digital advertisers to analyze data across Google Cloud, Amazon Web Services (AWS), and Azure. This solution is cost-effective as users don’t have to leave the BigQuery user interface.

BigQuery Omni is the first usage of Anthos by Google, which was rolled out last year. Anthos makes it possible to package your service and run it seamlessly on any cloud. So this means that Google is using Anthos to run BigQuery on top of AWS machines. Now that BigQuery Omni is leveraging Anthos—we’re one step closer to cloud

While the industry continues to develop solutions leveraging cloud commoditization there are still many questions remaining around governance, complexity for users, and who will own the service level agreements (SLA).

These recent changes will enable digital advertisers to use a single tool to examine data instead of manually moving data from place to place. Whether this solution is interesting for an average advertiser will depend on the additional costs incurred versus the volumes of data involved.

Cost & efficiency 

There are a couple of key challenges BigQuery Omni tackles that have been impacting digital advertisers for years. The first is the high-cost associated with moving data between cloud providers.

When BigQuery was introduced, it created a process for data to be accessible through a data warehouse format meaning data could be scaled quickly and at a low cost.

Now BigQuery Omni represents the same way of analyzing data stored among public clouds.

To best explain this particular process—suppose data is stored in AWS and an advertiser wants to run a data analysis.

Before Omni, advertisers had to design, develop and operate a system to run data imports on a regular basis, check the data transition consistency and schedule updates to be in sync across the two repositories.

This drawn-out process created unnecessary costs, including maintaining and monitoring the processes over time as well as the network transfer costs as the volume of data leaving a cloud is usually charged on a per-volume basis.

With BigQuery Omni, the consistency is already there, thus eliminating additional development cost. Network transfer costs are also cut as only the query result goes over the network and query results are usually smaller than the whole set of data processed.

Data management

Handling data is more efficient with the new BigQuery interface. Advertisers can access data stored in Google Cloud, AWS and Azure through Google BigQuery eliminating the need for copying data or moving from other cloud solutions. Users can save results from BigQuery Omni within the platform resulting in no cross-cloud movement. 

When companies contact a cloud provider, they go in-depth to analyze the guarantees around security, privacy, compliance and continuity of service. Things get more complex when you want guarantees on a service that spans over different providers.

For instance, when a company wants to conduct security audits or have a look at certifications, the information may be difficult to find in multiple clouds. With BigQuery Omni a significant amount of the requirements are on AWS or in GCP or might be in BigQuery Omni.

Chances are BigQuery Omni users have contracts already in place with each cloud provider. How will this hybrid setup fit and who shall advertisers ask to get a full SOC 2 report of this hybrid service? What can be expected in terms of processing rates considering negotiated specific rebates?

Companies will want to have a look at security audits and other certifications. When considering using a multi-cloud service, know where the contract is sitting. When contracting with BigQuery Omni a big portion of the requirements are on AWS, as a customer, chances are users will not access all the details directly.

Governance

BigQuery Omni is already offering up ways to ease the complexity for users when it comes to governance. By using standard query language (SQL) and the same BigQuery API’s, users are able to break down data silos allowing them to analyze insights for their business in one place.

Unfortunately, this is just the beginning of the solution.

If a company’s data is held in different clouds, the company is more likely to have a different team for each cloud. Each team probably has its own set of practices and conventions that may not necessarily be consistent.

Opening datasets to Omni still requires an alignment in terms of exposed metadata and documentation for users that are not familiar with a part of the tools and processes.

Overall, cloud solutions like BigQuery Omni from Google are geared at helping BigQuery users access and securely analyze data hosted in AWS and will soon include Azure. This is one more sign of the cloud provider competition moving away from the raw performance battle to the ease of use and high-level feature comparison, at least for the most common needs.