11 Tips For Successful Self-Service BI And Analytics
As more businesses attempt to compete with data, more people within their organizations must be able to gain insight from it. End-user requirements are changing rapidly, often at a faster pace than their employers' ability to deliver sound solutions. Here are a few ways to avoid compromising long-term benefits for short-term gains.
![](https://eu-images.contentstack.com/v3/assets/blt69509c9116440be8/blt188e1a60ae9bd93b/64cb3845ae8f30e47ebee4e6/analytics-925379_1280.jpg?width=700&auto=webp&quality=80&disable=upscale)
Businesses competing on data must be masters of change. To keep pace with constantly shifting business models, markets, and customer expectations, companies must become more agile, which includes empowering employees with insights that are available at their fingertips.
Self-service analytics is one of the tactics separating industry leaders from laggards.
In today's world, "self-service" is no longer synonymous with passively consuming static reports pre-packaged by IT. It's more about building one's own reports, exploring data, and interacting with it.
Self-service BI and analytics solutions are continuing to evolve to meet the requirements of agile, data-driven enterprises. As a result, the vendor landscape is changing radically. So radically, in fact, that Gartner reimagined its BI and analytics Magic Quadrant for 2016.
Create a culture where technology advances truly empower your business. Attend the Leadership Track at Interop Las Vegas, May 2-6. Register now!
By 2018, Gartner expects self-service to mean more than analytics. By then, most business users and analysts will have access to self-service tools that also enable them to prepare data for analysis.
While software can abstract the underlying complexity of connecting to data sources or executing an analysis, software is not a complete substitute for human intellect. Although predictive and prescriptive analytics can help, humans still need to understand how they can use data to benefit the business. In addition, specialized knowledge is still necessary to solve unusually difficult problems.
Wherever a company is on its journey, there is no shortage of tools from which to choose. But even with "the right" tools in place (which varies from organization to organization), businesses can still fail to realize the potential business value of their efforts because they still have organizational obstacles to overcome, not the least of which is balancing the pace of business and technological innovation with a corporate structure and mindset that can support it.
Here are some ways to realize longer-term value from self-service BI and analytics investments.
Culture is one of the biggest factors separating data-driven companies from their traditional counterparts. However, having self-service capabilities in place is only valuable if people in the organization are using them. At Ace Hardware, the directive to use data in everyday decision-making comes from the top.
"Years ago, [self-service capabilities] were in pockets in the company. Now we have more of an emphasis from our leadership that there are numbers out there and we can be a better company if we're using them, so we have adoption pretty much across the board," said Tim Brynda, a software engineering consultant at Ace Hardware, in an interview.
In 2014, Ace Hardware employees consumed 162 billion lines of report data. Apparently, people in the company disdain top-level numbers and executive summaries. They want detailed data, Brynda said.
The rapid pace of business and multi-year enterprise system rollouts are at odds with each other. To avoid spending millions of dollars and years of effort on something that may be outdated before it launches, more companies are phasing rollouts or adopting agile practices that are narrower in scope. That way, they can start small, learn from their mistakes, address issues proactively, and build upon their successes.
"You have to understand and prioritize the business problems, and you need a transparent cross-organizational process for buying into what you're trying to achieve and how. You also need to have an eye toward the future," said Mark Heslop, vice president of product management at pricing and profitability solutions provider Nomis Solutions, in an interview. "Having a more agile approach is key."
While understanding user requirements has always been important for the successful deployment of any technology, users tend to have very definite opinions about what they expect from self-service analytics -- their level of sophistication, ease of use, ability to provide insight quickly, effectiveness of data visualizations, and more.
Companies should have a clear vision of what they want to achieve with self-service BI or analytics, but beware of thinking too narrowly. One company hosted a competition to see which of three departments could drive the most value out of self-service analytics -- sales, marketing, or finance. The HR department, which had been excluded, "won" by gaining access to the system and analyzing employee-related data no one else had considered.
"No one gave them access, but they won the competition by realizing the greatest amount of business value," said Amir Orad, CEO of BI and analytics solution provider Sisense, in an interview. "Don't try to limit access to self-service analytics based on a certain title, department, or power users."
Understanding what data says is one thing. Taking sound action on it -- or even knowing what action to take -- can mean the difference between tangible and dubious ROI.
Different users have different requirements, which is why there are so many choices available for experts, novices, and everyone in between. Role-based capabilities recognize the need for varying levels of information abstraction. However, they may fail to consider the capabilities of individual users. Recognizing this, Ace Hardware implemented three versions of self-service with InformationBuilders' help. Now average users can build and save reports using a wizard. Those with knowledge of databases and fields can use a drag-and-drop virtual assistant. The most adept can use basic templates to create their own reports.
"We're doing a lot more fact-based decision-making, so, if we have a sale, did sales increase that weekend or did people just come in and buy that one item? If a hurricane is coming, what items should we stock? We're able to make better business decisions," said Tim Brynda.
Just because analytics purchases don't always flow through IT doesn't mean that data governance isn't still important. Without solid business-IT coordination, the business may make purchases that adversely affect infrastructure. In addition, the adoption of point solutions can lead to more information silos that can obstruct the flow of important information.
Somewhere between an IT centricity that is too slow to keep up with the pace of business and unbridled shadow IT is a balance of expertise and resources that enables the business to get faster insights to data without exposing the company to unnecessary risks.
"You want to make sure that available data structure can be used by any tool you have, so you're not building proprietary data systems," said Francois Ajenstat, VP of product development at BI and analytics solution provider Tableau. "You should have a governed infrastructure that leverages the security you have in place, provides the performance users need, and provides the ability for people to answer their own questions."
A lot of time is wasted within and among various parts of an organization when people are solving the same problems time and again. Self-service BI and analytics don't inherently help. Best practices do.
"People buy point solutions because they have pain points, but they're not thinking about how to operationalize the data so that the answers are available to more people," said Timothy Alexander, senior product marketing manager at fast data solution provider TIBCO Software, in an interview. "People run an analysis that sits on a random server that only X people can access, so when the question comes up again, they'll look for a new way to answer the question. We don't have to keep making the same effort on the same questions."
What works well on a small scale isn't necessarily sustainable on a larger scale. Facebook found that out as more of its employees started using self-service analytics. In 2007, the company relied on a massive piece of infrastructure like many other companies, but ultimately scalability and quality of service became issues. So, the company became an early adopter of Hadoop.
"When people talk about self-service analytics, they forget about the infrastructure part. [By] 2011, use-cases started sprouting, because the infrastructure and the tooling around it [were] made into a self-service platform for big data. The whole thing was very transformative," said Ashish Thosoo, former head of big data at Facebook and CEO and founder of data-as-a-service company Qubole, in an interview.
By 2011, after a four-year effort, 30% of Facebook employees across all departments were using the self-service analytics capabilities -- including developers, advertising operations, legal, product management, user operations, and security. The overall goal was to fuel the company's rapid growth, which, at the time, included a target of 1 billion users. The company has since exceeded that goal by more than 50%.
-
About the Author(s)
You May Also Like