Google unveils new Looker tool, BigQuery pricing models | TechTarget

• Read original article here

Google on Wednesday unveiled Looker Modeler, a new standalone service that will enable customers of any BI vendor to use Looker's semantic modeling capabilities to define and store metrics.

The service will be available in private preview during the second quarter of this year, with general availability scheduled for later in 2023.

Google acquired Looker for $2.6 billion in 2019 to add semantic modeling and business intelligence capabilities to its data cloud platform.

Since then, the cloud computing giant has integrated Looker with other Google tools, and last year consolidated its other BI tools under the Looker name.

Now, Google is developing Modeler, a service that breaks out Looker's semantic modeling capabilities from the rest of the Looker platform and makes them compatible with other analytics platforms.

In addition, Google introduced new pricing editions for BigQuery, the vendor's fully managed cloud data warehouse, and a tool within BigQuery called data clean rooms that enables users to combine their own data with third-party datasets.

The new BigQuery pricing editions are generally available, while data clean rooms will be available in preview during the third quarter of this year.

When Looker first launched its analytics platform 10 years ago, one of the ways it sought to stand apart from competition including Tableau and Qlik was by including semantic modeling capabilities. Semantic layers enable administrators to define key metrics and standardize the meanings of terms across their organizations to create data consistency and avoid data duplication. That subsequently enables self-service users to find and share data without having to know code or how to query and join multiple tables or sources. Recently, Looker has developed integrations with other analytics platforms, including Tableau and Google's own Connected Sheets. And on Wednesday, Looker added an integration with independent analytics vendor ThoughtSpot. Through those integrations, users of those non-Looker platforms have access to Looker's semantic modeling capabilities. Looker Modeler will make those semantic modeling capabilities available as a standalone service so that users of any BI platform can take advantage of the technology. Pricing details will be available once Modeler is generally available. It's that compatibility with third-party platforms that makes Looker Modeler a significant new feature, according to Doug Henschen, an analyst at Constellation Research. "[Modeler] exploits the product's core strengths while opening up use with third-party products," Henschen said. "Google made obvious moves to add integrations with BigQuery and other Google Data Cloud services, but I like the fact that it has maintained Looker's compatibility with myriad third-party cloud data platforms." Similarly, Mike Leone, an analyst at TechTarget's Enterprise Strategy Group, noted that the openness Google is demonstrating by making Looker's semantic modeling capabilities available to users of any BI platform is significant. Vendor lock-in is a concern as organizations develop their data pipelines and analytics stacks, and Looker Modeler enables organizations to choose the tools that best fit their needs. "Openness has always been a pillar to Google Cloud's message, and the interoperability of Looker Modeler continues down that path with support for all the major BI tools," he said. "Whether using Looker for BI and data visualization or another tool, Looker Modeler can serve as the data foundation." In addition, Leone noted that data modeling is crucial to ensuring data quality and enabling collaborative decision-making. Looker Modeler opens those modeling capabilities to any organization without forcing them to subscribe to the entire Looker platform. "Data modeling plays a critical role in ensuring that data is effectively used and shared across the business," he said. "Together with governance, people are enabled to have access to the right information they need for their jobs. When data modeling is done right, you'll see a significant boost to data quality." Google developed Looker Modeler at the request of customers, according to Gerrit Kazmaier, Google's VP and GM for data and analytics and Looker. While some customers of Google's myriad data cloud platform tools are also users of Looker, many others use other vendors for their analytics needs. And like Looker customers -- and users of platforms with integrations with Looker -- they want access to Looker's semantic modeling capabilities, according to Kazmaier. "We have heard consistently from our customers that getting their [metrics] consistent is a struggle," he said. "It's difficult to align numbers from meeting to meeting, and they truly saw the benefit of Looker having trusted and governed data. Now, with Looker Modeler, customers can get the modeling service only, and connect it to the BI tools of their choice."

Beyond the introduction of Looker Modeler, Google on Wednesday unveiled BigQuery data clean rooms in a move designed to help organizations securely improve their marketing efforts. With the capability, users will be able to combine their own marketing data with third-party data through the Google Cloud data marketplace to develop a fuller and deeper understanding of how to target potential customers, the vendor said. And given BigQuery's governance capabilities, they'll be able to combine their internal data with external data securely to protect the privacy of data and ensure organizations remain regulatory compliant. In addition, customers that choose to will be able build their own data clean rooms on BigQuery with their own specific data governance frameworks rather than use the data clean rooms provided by Google, Kazmeier said. "They are basically our technology … to make sure we have strong data privacy [measures] and allow our customers to use that technology to upload their first-party data into BigQuery and analyze and combine it with other datasets in a privacy safe way," he said. Henschen noted that other data cloud vendors including Databricks and Snowflake have introduced similar features and BigQuery data clean rooms is Google's response. Leone, meanwhile, pointed out that trusted data environments are crucial, which is why Google and some of its competitors are now taking steps to provide customers with data clean rooms. "It's a big reason why we've seen several clean room announcements over the last several months," he said. "This is especially important in highly regulated industries like healthcare and finance that have multiple parties across their business working together but must protect individual privacy and maintain compliance." Google also launched a new pricing structure for BigQuery aimed at helping organizations better predict their cloud computing costs. Previously, Google offered only on-demand analysis pricing at $5 per terabyte for queries and flat-rate analysis pricing at a monthly cost of $2,000 per virtual CPU (which Google calls slots). Now, with the introduction of BigQuery editions, customers can choose from three consumption-based pricing options -- Standard, Enterprise and Enterprise Plus -- that Google said have the potential to result in more predictable pricing. Standard costs $.04 per slot hour and is optimized for standard SQL analysis, Enterprise increases the cost to $.06 per slot hour and is designed for more advanced analytics, and Enterprise is priced at $.10 per slot hour and includes more application-specific capabilities such as FedRAMP compliance. Each option comes at a reduced rate with a one-year commitment. In concert with the introduction of new pricing options, Google said that as of July 5, 2023, flat-rate slot commitments will no longer be an option and the cost of on-demand analysis will rise by 25%. "Google has continued to tweak its BigQuery feature set and pricing model to be more competitive in response to the market and customer requests," Henschen said. "Cost optimization is very much on the minds of customers these days. Having more options is better, though it's typical to see the steepest discounts tied to long-term commitments." While Looker's semantic layer continues to differentiate it from its peers, the platform's augmented analytics capabilities fall short of those from its competitors, according to Henschen. Ask Looker, a natural language query tool, is in development. But many vendors, including Tableau and ThoughtSpot, already offer NLQ tools of their own. Meanwhile, as Looker develops integrations with other Google tools that have the potential to add augmented analytics capabilities, it needs to keep in mind that not all Looker customers are also customers of Google's other tools, Henschen said. For example, Looker can take advantage of BigQuery AutoML through BigQuery's ML Accelerator. But rather than integrate with a host of other augmented intelligence and machine learning tools from Google -- which aren't available to Looker customers that don't also subscribe to other Google products -- Looker would be better served by developing some of its own AI and ML tools. "Augmented capabilities continue to be a work in progress," Henschen said. Eric Avidon is a senior news writer for TechTarget Editorial and a journalist with more than 25 years of experience. He covers analytics and data management.