At Astreya, we are proud of all the services we provide, from managing IT support for end-user devices to the design, selection, and implementation of cloud infrastructure. However, we cover so much territory that sometimes people do not realize just how strong we are in business intelligence (BI). Having managed BI programs for a multitude of companies in the Fortune 50 for several years now, we have developed unique expertise in conducting data management and analytics at enterprise scale. We believe our skills in deploying agile BI and BI automation programs are unmatched and our clients’ satisfaction is an endorsement of this belief.   

In this post, we will describe how we expanded from having just a handful of consultants on-site at one of the largest social media and technology companies in the world to having over 120 consultants on-site daily with ~150 more ramping up in the next quarter —  we accomplished this in less than three years. We are proud of that level of growth because it only comes when you can consistently deliver value across a broad spectrum of services and earn the trust of the client. I recently spoke to Astreya business operations manager Minh Phan, who is leading our BI efforts with this customer to understand how we built this level of trust so quickly and with such a large and complex client.  

Before we get into how Minh succeeded in becoming a critical part of this client’s overall BI initiative, let’s look at why most analytics initiatives in large enterprises fail. Companies may appear high tech from the outside but they are loaded with data silos and they are far from being able to implement analytics consistently across all corners of their enterprise. In fact, billions of dollars are poured into BI every year yet more than half of all analytics projects fail according to Gartner. This typically happens for two reasons: failure in the first mile or the last mile of the analytic journey. 

  • Failure in the first mile: Eager to show results, many neglect the difficult and less attractive work of data management required to clean data and gather it from across the many siloes it resides in. Data management and data quality are essential first steps in making sure that the analytics dashboards and reports are trustworthy and aligned with business goals.
  • Failure in the last mile: Most BI programs fail to solve the last-mile analytics problem. They stop at the creation of dashboards that require monitoring and interpretation in order to add business value. What people really want and need is actionable data-driven insights that can help to grow the business. Don’t just visualize the data, interpret it and distill it down to steps that should be taken as a result of the insights gained.

To overcome these challenges, Minh and his team were obsessive about data management from the very start. Often, BI professionals race toward dashboard creation and spend much of their time and energy creating really attractive looking dashboards. But if you distribute dashboards based on bad data, you are doing more damage than good. More important than quickly developing something that “looks good” was developing a deep understanding of the business needs and objectives of the client with quality data to develop something that “is good”.

With these guiding principles, Minh’s team hit the ground running by gathering as much data as they could. They established a culture of continuous improvement and created new data sources when necessary, and in many cases, added value by enriching existing data sources. They made it a point to become experts in the client’s existing homegrown BI solutions, becoming power users. Soon they knew as much about these internal tools as anyone on the client-side and made insightful recommendations to improve those tools. Minh adds that we are very careful not to slow down the client during this process and run alongside them in parallel. We pride ourselves on moving as fast or faster than our clients and providing a fresh set of eyes on their infrastructure to find a way to automate, innovate, scale, and improve efficiency.

From the beginning, Minh’s goal was to make as few dashboards as possible. His guiding question, “How do I build fewer dashboards but increase the confidence and usable insights?” This led him to explore BI automation tools for dynamic insight delivery. Why wait and hope that people will regularly look at your dashboards and then be able to extract the right insights when you can set up automation to deliver the insights directly? One of the more promising tools Minh came across for BI automation was RelayiQ. RelayiQ empowers you to:

  • Detect changes in KPIs based on your data using machine learning
  • Prescribe customized action plans based on those changes and insights
  • Notify those who can act on them in a timely manner

Shortly after selecting RelayiQ as the right vendor for his client’s needs, Minh learned of Astreya’s plans to acquire RelayiQ. It was a perfect fit. Astreya’s BI experts like Minh and his team know how to solve the arduous first-mile analytics problem of data management. While this needs to be a prerequisite to any analytics program, it is often neglected — hence the high failure rate of BI project deployments. RelayiQ now adds to that a solution for the last-mile problem in analytics by automatically detecting important changes in data and immediately alerting all the right people to act on it with a clear set of instructions.

While it is early in the implementation of the BI automation we are well-positioned to take advantage of it precisely because of the early investments made in data management. Those investments and staying outcome and value-driven are also paying off through:

  • Introduction of data into the decision making cycles across all levels of the organization
  • 75% less time spent preparing business review decks each month and more time spent actively solving issues, innovation
  • Effectively absorbing the increased support responsibilities resulting from double-digit % network expansion while maintaining flat operating cost with no impact on service quality

Do you feel like your large enterprise may never be able to get its data in good enough shape to run trustworthy analytical reports? Do people largely ignore and distrust the proliferation of dashboards across your organization? Would you rather spend timely insights with the corresponding actions than gamble on people extracting insights themselves from cryptic dashboards with suspect data? If you answered yes to any of these questions, we would love to engage and share more with you about how we have delivered on the true promise of big data, BI, and data science for many of the largest and best-known companies in the world.