Whether you call it self-service analytics or self-service business intelligence (BI), there has been much discussion about the perils, myths, promises, and prospects of successfully building self-service capability. Going forward, I’ll use the phrase “self-service BI” but you are welcome to substitute the words “self-service analytics”.
So, is self-service BI actually attainable or just snake oil? If it is attainable, then what do you need to know to ensure your self-service BI endeavors are successful? In this article, I’ll walk you through the key benefits of self-service BI and some of the top concerns. I’ll also share what you need to do to build a successful self-service BI program.
What exactly is self-service BI?
Self-service BI allows business users to ask questions of their data, get insights, and analyze data—without relying on IT, BI specialists, or SQL for the purposes of taking business actions. Now, let’s break this down just a little further:
While self-service BI doesn’t require specialized technical expertise, Self-service BI does rely on the knowledge and expertise of business users to know which questions to ask next and what actions to take based on the answers.
The cycle of insanity
Most data engineers and analysts are familiar with the typical data analysis request that I call the “cycle of insanity”. A stakeholder makes a request for dashboards or to have questions answered, followed by the realization that the data doesn’t exist in the system, data engineering needs to ingest, secure and govern the data, and the analyst builds out the dashboard, only to have the request refined to answer the next question.
While this approach may provide value to the business and the data teams may feel valued, the problem is scale. This process does not scale, not even for small companies. The business wants visibility into its operations, to intuitively use the data they have, and to focus on the business operations. Analysts overwhelmingly want to provide insights that drive value for the business and not on mundane joyless data extracts or more deadboards.
When Frontify realized this challenge in their own organization, there were only three people supporting the BI insights for 200 employees. The data team painstakenly pulled data from business applications into a MySQL database using custom Python scripts, and then analyzed it in Tableau.
“It was a real pain point. We were stuck in a loop of constantly updating reports.”
It could take up to a month to produce reports and dashboards, by which time the underlying data was often obsolete or the opportunity for actionable insights was missed. They knew they needed to scale their analytics program to improve metrics across the company—that’s when they started looking for self-service BI solutions.
Data volumes are growing, business leaders are demanding more insights, and each new question yields additional questions. This, in turn, creates more and more data requests from data teams, which is simply unsustainable. That’s where self-service BI finds its sweet spot—backlog reduction—freeing data engineers, analysts, and business users from the never-ending cycle of insanity and dashboard hell.
Benefits and historical concerns of self-service BI
Let’s take a look at a few of the primary benefits of self-service BI and the chief concerns associated with them. After that, we’ll look at a few things to consider when building out your self-service BI capabilities.
Scaling your BI capabilities
Self-service BI solves the problem of scaling your business intelligence capability. Enabling users to answer their own questions, ask the next question, get their own insights, and take action, does a few key things.
Empowers business users.
Reduces the backlog of data teams and enables them to focus on truly complex business problems.
Leverages the existing security and governance policies that exist within the organization.
Historical concern: Empowering business users that may not have the skills or even desire to take on such tasks, may reduce the value of self-service BI.
“Passion, when it comes to data, [data] literacy is really important. That's why we put the GM analytics academy course out there, to not only build the literacy itself, but just the passion in having a core competency around analyzing and presenting data in a compelling way to make decisions.”
Iwao Fusillo, an esteemed Data Analytics Executive at PepsiCo, understands the importance of data literacy and fluency. During his time as GM’s Chief Data and Analytics Officer, Iwao joined Cindi Howson on The Data Chief to explain how he analyzes talent and implements BI and analytics programs that fuel adoption. Listen to the episode.
Reducing the licensing, development, and operating costs
Identifying and realizing cost reduction opportunities is often a critical factor in getting leadership support for new programs and initiatives. Here are three areas where self-service BI can help you realize actual savings for your organization:
Licensing cost - A well-constructed modern data stack with consumption-based solutions ensures that you are paying only for what you are using. Traditionally, BI tools are based on user licenses so moving to consumption-based tools enables you to stop paying for unused licenses and only pay for what you use.
Reducing the backlog of mundane data requests to data teams, shortens delivery times, and enables highly skilled and compensated analytics engineers to work on initiatives that provide the most business value.
Being on a “mission to decommission,” redundant or legacy business intelligence tools that don’t advance your goal of self-service BI are targets for cost savings; not just licensing costs but lost productivity, ongoing maintenance, and support cost.
Historical concern: Integrating and securing new technologies may add cost and delay the time to value of self-service BI.
Read how Schneider Electric’s End To End Data Flow Architect and former People Analytics Change Manager, Alanna Roesler achieved an incredible 78% ThoughtSpot adoption rate, allowing her to re-allocate an impressive 25 hours per week towards driving digital transformation efforts.
Increasing your speed to insights and actions
Forward-thinking business and data leaders know that futurist Jim Carroll is right, when he declares, “the future belongs to those who are fast.” Markets and industries are changing and are reacting to market forces faster than ever. For example, when the pandemic disrupted supply chains and there was uncertainty in the auto finance industry, auto lenders, wholesalers, and dealers that pivoted quickly to the pre-owned car market actually increased their revenue and profits while laggards missed the opportunity. This illustrates two of the major benefits of self-service BI:
Empowering business and non-technical users to get insights quicker and take action.
Enabling business users to create more value for their customers.
Historical concerns: Users may not fully understand the data or know how to make inferences from the data.
In a recent episode of The Data Chief, Vijay Kotu, SVP of Data and Analytics for ServiceNow, discusses the importance of empowering frontline workers to make data-driven decisions, explaining that these decisions can’t just happen once a quarter in a boardroom.
What you need to consider when building your self-service BI capability
You have seen the primary benefits of self-service BI; you are ready to embark on a self-service BI journey. To ensure the success of your self-service BI program, do these three things:
Think big, start small, and scale fast
Think big about what you want your organization to be – it’s a must. Start small by picking use cases that provide quick wins and business value. They also build momentum. And scale fast by showing these quick wins across the organization.
Business leaders are competitive; it’s human nature. We all want the best for our customers and employees. So, seeing how co-workers are winning with self-service BI will drive adoption and increased funding across the organization. While business users may initially understand how self-service BI fits into their processes, seeing their co-workers winning will be a huge motivating factor.
Forge culture of data-driven, self-service BI is a cultural shift
Building a data-driven culture is part of any self-service BI capability and will require an intentional effort for improvement, an analytics champion, and support from senior leadership. According to New Vantage Partners, 92% of organizations attribute the “principal challenge to becoming data-driven” to people, business processes, and culture—only 8% identifying technology.
For your self-service BI program to be successful, consider building a data literacy and enablement program. Regularly scheduled “Office Hours” are a common practice that solves problems and builds trust. Gamification is also employed, including:
User-led presentations demonstrating “I built this”
Awards and badges for milestones, such as “1st Liveboard” or “10th Insight” generated
Leaderboard for answers and liveboards.
All of these create a sense of trust, support, and community within the organization.
Elevate data management practices for well-governed and well-modeled data
The analyst of the future is slightly different from the analyst that sits in the center of the cycle of insanity. Future analysts will optimize analytic data and data models to enable self-service BI and enterprise scalability. They will focus on the most challenging analysis, not the most mundane. While most of the tools and skills remain the same, optimizing for scale and self-service BI will be an additional skill.
To fully unlock the potential of self-service BI, analysts will need to create data models that answer questions for a larger domain rather than ad-hoc reports or dead dashboards. Data models in self-service BI will dovetail with existing governance, security, and business logic to ensure that non-technical business users have the right data to make the right decisions.
Self-service BI is not snake oil, and it can be implemented by organizations that embrace the modern data stack. Yes, it’s attainable, but like anything of value, it’s worth a bit of planning, organization, and process. Your BI capabilities will depend more heavily on the people and process than it does on the technology. However, if the technology for your data experience layer doesn’t support augmented intelligence, search-based analytics, and AI-driven insights, you’ll have a hard time achieving the benefits listed above. As for the people and process components, focus on building a data-driven culture for your organization, and remember that well-governed and well-modeled data will enable scale within your lines of business.
So next time you or your team is caught in the cycle of insanity, rest assured that breaking the cycle is possible and let your business stakeholder know there is a way out!