Given the possibilities for innovative, new solutions in today’s digital world, Accenture has undertaken the goal of transforming into an insights-driven enterprise that embeds analytics into the core of its operating model.
Accenture used to operate on siloed data, pervasive use of spreadsheets, limited sophistication and multiple versions of data. That’s why our teams came together to define a cutting-edge analytics platform and capability that could deliver actionable, data-driven intelligence across our business. The result? More than 20,000 colleagues and counting using more than 120 analytics products in excess of 100,000 times a month. Fifty percent of legacy reports retired. Transformation of our reporting landscape, establishment of an advanced analytics capability and more intelligent decisions driven by confidence in our data. And a business that’s ready to excel in the new, and help our clients do the same.
Companies and industries around the world are pivoting to the New. They’re discovering and exploiting cutting-edge technologies. They’re finding new ways to meet the ever-evolving expectations of their customers. They’re seeing old models fall away and new ones emerge. But it creates a challenge for a company like ours: How do we lead in a game without any fixed rules?
become an insights-driven enterprise
One characterized by automation, optimization, prediction and continuous learning and, of course, a data-native culture. That creates its own imperative: Insights-driven businesses embed data and analytics at their core. But we’re an established company with hundreds of thousands of employees spanning the globe. And, like any other, we had our share of siloed data, duplication and value trapped in spreadsheets. But we knew it wasn’t good enough for an insights-led business that aspires to lead in the new. We needed an analytics platform and enterprise-wide capability that could evolve at the pace of change. Or even outpace it.
Teams from across Accenture came together to define a solution. We had to find a way of managing our data at enterprise scale, while supporting the whole spectrum of analytics needs across our organization. And we had to ensure they were driving real business outcomes.
First, the data. How could we turn hundreds of sources and thousands of reports into a common platform? It’s not a problem with a single solution.
Instead, we asked: what are our objectives? What are the common patterns? How can we reuse them? An example: looking at the different ways data is ingested revealed just two simple underlying patterns: constant streaming and batch processing. Focusing on these patterns, not the many different use cases they support, meant the effort to complete the design and analysis could effectively be commoditized. Even better, it creates a framework for switching technologies in and out as needed. That’s an incredibly agile and cost-effective way to manage data.
Then, the analytics.
We needed to support the entire spectrum of analytics needs, everything from conventional business intelligence right through to flexible and speculative work in data science and machine learning. Again, we looked at business outcomes: reporting (what you know); analytics (what you know you don’t know); and data science (what you don’t know you don’t know).
For BI and reporting, we created “nearshore data marts”—small, manageable data sets serving a specific need—that are quick to develop, easy to change and housed on fast-access systems such as SAP HANA®.
For data analysts, we went further. Centralizing data security, privacy and management, meant we could give them fast self-service access to a secure environment for generating insights across broad swathes of previously inaccessible Accenture data. And new tools let them blend data sets into visualizations and immediately share the results.
For our data scientists, we went further still. Using next-generation tools and agile, open-source development, we created advanced forms of self-service data science model management.
With a single click, data scientists can now create models for use across multiple applications. What’s more, they can generate faster intelligence by collaborating, verifying and improving each other’s results.
Finally, we wanted to accelerate the use of advanced analytics across our business. So we created the Analytics Studio.
Staffed by data scientists, it’s a lightweight, push–pull innovation model. Using a time-boxed, fail-fast mindset, it lets our teams explore using data to change the way Accenture does business and push out minimum viable products at pace. And it’s centrally funded. So no one wastes time sourcing money to explore new ideas.
Our transition to an insights-led enterprise is advancing at pace. More than 20,000 employees have already shifted from static, asynchronous reports and spreadsheets to interactive, online, near real-time operation. They’re using more than 120 enterprise-wide analytics products to interact with our data in excess of 100,000 times a month. And they’re getting new, actionable intelligence on everything from clients and sales to human resources and global management.
We’ve also eliminated 90 percent of the repetitive, time-consuming steps previously needed to create new analytics and AI models. That’s enabling our data scientists to answer up to 50 queries a month from the business. They’re now providing new intelligence in days or hours rather than months. And they’re finding the big wins we can quickly scale across the business.
In creating a cutting-edge analytics capability for our entire business, we made the decision to lead by example. And we’re showing just what an insights-led enterprise can achieve.