In its battle for public cloud supremacy, Microsoft announced a new service today that it plans to debut next month: A predictive analytics and machine learning tool.
The move is interesting on a number of fronts. Firstly, it shows that the company is putting the peddle to the metal in terms of rolling out new products, services and features in its cloud computing portfolio to take on the giant of this industry and its biggest competitor - Amazon Web Services.
Secondly, focusing on data analytics is a natural area for Microsoft to invest in. The company has a strong history of managing data for its customers. Between SQL Server and its suite of Office 365 apps, Microsoft a lot of corporate data is managed using software from the Redmond company. In a new age of next-generation data management, Microsoft is a natural company to provide analytical features from its cloud. (Read about the Machine Learning news announcement at Microsoft's website here.)
But perhaps the most interesting part of this news is how Microsoft is taking a queue from AWS's own playbook and focusing on data analytics. It's a smart and savvy move, and customers should be aware of the dangers of placing a lot of data in the cloud by remembering one thing: Data is easy to put in the cloud, but can be tough to get out.
Check out a video of Microsoft announcing the service here:
The reason users should be cautious about putting lots of data in the cloud is because data has what is called "gravity" with it. Once data - particularly large amounts of it - is stored someplace, it's hard to move it. If customers store data in Microsoft's Azure cloud, then it gains perpetual revenue for as long as that data is stored in its cloud. For that reason, big public cloud companies like AWS, Microsoft and Google want as much customer data as possible in their clouds. And what better way to attract companies to store data in your cloud than by offering fancy machine learning and analytics services on that data?
AWS has taken the same strategy. Two years ago at the company's first-ever user conference the company announced RedShift, a data warehousing product that companies can use to store data and then provide analytics on top of it. Between Simple Storage Service (S3), its range of databases - DynamoDB NoSQL, Elastic Map Reduce Hadoop-like processing and Glacier cold storage - Amazon has made a strong play to be the preeminent storage spot for data in the public cloud.
It's a savvy strategy for the big cloud players. The more data they get, the bigger their clouds become and the more recurring revenue they get. They know that once customers put data in their clouds, they will likely keep putting it there. And they also know that it's easy to put data into the cloud, but it's much harder to get it out.
Now there are some attractive benefits for users to keep their data in the cloud. Namely, they don't have to buy expensive hardware to keep it on their own premises. Micrososft, Amazon and other cloud providers will store it and if you need to grow your capacity, it's just another charge on the credit card.
The economics of the cloud can be compelling and these analytics services and products can be great. Microsoft says its Machine Learning capability can help financial institutions detect fraud, help health care providers give more efficient care and help retailers make smarter decisions about how to stock their shelves. That's all well and good. But all of those decisions take a lot of data to analyze and provide actionable insights on. Just remember that once that data goes into the cloud, it can be tough to get it out, and you'll be paying to keep it there.
This story, "The important thing to keep in mind about Microsoft's new Machine Learning cloud tool" was originally published by Network World.