by Alan Herrera, Sr. Product Marketing Manager for Advanced Analytics
In recent years, the promise and, in some cases, the hype of big data have produced plenty of critics who doubt the benefits of big data or point to the potential pitfalls organizations face when engaging in big data and analytics projects. Although these critics make some valid points (e.g., not every organization has big data or needs to manage and analyze big data), a more pragmatic approach is to look not just at big data but also small data, long data, dark data, and any other metaphor for real data — in other words, all data.
Why should we care about data? IDC recently conducted research commissioned by Microsoft that showed better outcomes from data and analytics projects correlating with greater competitiveness of an organization in its industry or a better ability to fulfill its mission in the public sector. Although correlation does not equate to causation, a growing body of research shows financial and productivity benefits directly linked to better data-driven decision making enabled by business analytics solutions. IDC’s research found that companies taking advantage of their data had the potential to raise an additional $1.6 trillion in revenue by 2017. IDC called this the data dividend (source: IDC, April 2014).
One of the most effective ways an organization today can leverage their data is through machine learning. Machine learning enables business to move beyond the “what happened” of traditional BI to the “what will happen” insights enabled by predictive analytics. Predictive analytics, coupled with Azure Machine Learning (AML), can translate past data and human input into deep statistical analysis, potentially uncovering patterns in the data and insights on business drivers that will drive improvement in the underlying business. Machine learning enables businesses to drive tangible return on investment. With AML, our customers can create solutions across a number of scenarios, such as customer churn analysis, fraud detection, targeted advertising, predictive maintenance, and more.
Microsoft has been using machine learning internally since the founding of Microsoft Research in 1992. An early example is the use of machine learning to improve Hotmail spam detection. We then went on to optimize Kinect functionality for Xbox, and more recently used natural language algorithms with Skype Translator to enable real time translation.
With Azure Machine Learning, we now have the ability to offer these kinds of solutions to our customers at scale. Our vision with Azure Machine Learning is to make Advanced Analytics accessible to everyone. We provide a modeling experience that welcomes all skill levels. Data scientists can use trusted algorithms from Xbox and Bing without writing a single line of code, and more seasoned data scientists can mix and match with Python and R built in, or even drop in their custom code. The tool literally speaks their language. Then, we can deploy in minutes as a web service—the one-click deployment is what makes our product so unique.
AML users can both deploy their solution to the community through the in-product Gallery or learn best practices for their own solutions. They can take it a step further with global scale through the Azure Marketplace, where partners can brand and monetize their solutions and developers can access them with no data science skills needed—the data science is inside.
How to partner with Microsoft on Advanced Analytics
In the last year, our Data Science Team created a unique new program called the Advanced Analytics Hackathon. The program allowed our customers to leverage the skills of our best Data Scientists to develop Advanced Analytics solutions in a span of 2-3 days. Through this program, we have worked with many industry leaders to help create solutions that improve their businesses. With the proven success of this program and the upcoming release of Azure Machine Learning, Microsoft would like to invite our partners to join in on this exciting opportunity.
Extending the Advanced Analytics Hackathon program to partners
The goal of an Advanced Analytics Hackathon is to scope and deliver an Advanced Analytics solution in an accelerated time frame. Step one is to identify customers who have a well-defined business program and determine if the issue can be addressed with Advanced Analytics. This is followed by work to determine specific success metrics for the solution. These success metrics will help the customer understand the impact of the solution and will help the Microsoft partner to position the value of the solution in tangible terms.
Advanced Analytics Hackathon metrics fall into these three categories:
- Return on Investment: What is the value off success vs. doing nothing?
- Productivity Downtime: How much will the solution help reduce downtime costs?
- Customer Experience: Will the solution improve the customer experience and help retain or acquire customers?
The output from an Advanced Analytics Hackathon falls into two categories:
- An operational solution that will drive Azure consumption
- A working Proof of Concept (POC) that can be scaled out through a longer-term Systems Integrator engagement
If the Hackathon focuses on scoping a future SI engagement to develop a solution, but does not deliver a POC or working solution, it’s not really a Hackathon. Delivering a working POC or operational solution is a critical output from an Advanced Analytics Hackathon.
We have scoped the program to require a minimum time commitment from our partners on site with the customer, but any successful engagement will require time ahead of the onsite session to scope and plan for the hackathon, and then time to follow up after the hackathon.
- Program length: 2.5 to 3 weeks total engagement time
- 1 week of planning to identify and sample data sources and determine success metrics for the solution
- 2-3 days on-site with the customer to build a solution in partnership with the LOB leader who owns the solution and their data science team and/or IT support team
- 1 week post-engagement support to ensure operational solutions are working as expected, to provide any necessary follow up training and plan and scope any additional SI engagement necessary to operationalize a POC
Why should you participate in this program?
Microsoft is deeply committed to delivering technology that enables broad consumption of Advanced Analytics and the value this brings to our customers. With Azure Machine Learning, we will broaden accessibility to machine learning and in doing so, grow the addressable market for Advanced Analytics solutions.
Partners that join this program will have the ability to share in this growth with Microsoft and build strong practices focused on Advanced Analytics. Planned benefits include dedicated readiness for the program and the opportunity to receive leads directly from Microsoft, along with additional resources to help our partners drive successful customer engagements and grow their Advanced Analytics business. We are in the process of planning the readiness and onboarding activities for the program and will share full details in the coming weeks.
Here are the draft requirements and benefits of the program:
How to participate and timing
If you’re interested in the Advanced Analytics opportunity and the new US partner program outlined above, you can request to join the Advanced Analytics Partners group in the US Partner Community Yammer network.
We will be recruiting our first set of partners between now and mid-August, and initial readiness activities are planned for early September. Scaled delivery of hackathon engagements is planned to begin in mid-September. In the coming weeks, we will share additional content and the timing of our readiness activities for the program through the Yammer group.
Advanced Analytics at WPC 2015
If you’re attending WPC 2015 in Orlando, you’ll hear more about Advanced Analytics. Hear Judson Althoff talk about your opportunities in the Microsoft North America General Session, Selling Experiences to Win, on Tuesday, July 14 at 4:00PM, Room S320. And, add these sessions to your WPC Schedule in Connect:
- CE24 – The Microsoft data platform: the opportunity is bigger than you may think
- CE18 – How to build your practice on Advanced Analytics technologies to your WPC Schedule in Connect, and
Microsoft partners talk about their experience with Azure ML