Eliminating Barriers to an Insight-driven Business: Do’s and don’ts from an Analytics champion on ground for 2 decades
Updated on

Eliminating Barriers to an Insight-driven Business: Do’s and don’ts from an Analytics champion on ground for 2 decades

By Bhaskar Dey with contributions from Praveenkumar Sathyadev.

These are great times for business intelligence and analytics practice teams. We have never seen so many variables aligned to enable the opportunity for data- and insight-driven business. There are more data sources, greater computing power, cheaper storage, better information management, and more intuitive analytical technologies. Yet, in spite of available technologies and practices, many BI implementations do not deliver the desired or anticipated results. Organization and partners of all sizes suffer from countless oversights and poor judgment calls during planning, technology selection, and roll-out plans.

This article discusses certain best practices around BI which every company can follow. We have also tried to do a deep dive around some mistakes to avoid to be able to launch scalable, cost-effective and next-generation BI and analytics best practices.

Best Practices: Course5 Perspective

    # 1 Get organization-wide buy-in

    Disjointed BI practices and failed universal adoption is a quick path to BI failure. To maximize BI success, it is essential to get company-wide buy-in. It is in everyone’s best interest; in the end every department especially Sales, Marketing, Finance, and Management will benefit from BI. To that end, the right parties should be involved from the go. An important aspect which companies often miss is getting IT buy-in early. , As we are seeing these days, the CIO and CTO of the organization have a major say when it comes to any BI and analytics practice.

    At Course5 Intelligence, some of our leading tech and CPG customers have been able to drive business decisions with actionable data and insights – which initially came from first level BI dashboards.

    # 2 Start small

    Yes, every organization has lots of business questions that need to be answered on day 1. Is that really needed? Million-dollar question. When analytics teams and partners start compiling individual business questions and needs, the sheer size of the data required to be moved can be overwhelming. This is only compounded when every employee starts clamoring for data immediately. We recommend starting with a small list of crucial and mission-critical data questions. First try answering the questions with available data, and if that does not work, find out other departments that might have that data. If that fails as well, the time is ripe to explore data pipelines that need to be brought into the gambit of existing data sets – sales, marketing, transaction, clickstream, etc. However, do not try to do everything in one go, build small project plans and execute towards it. Gather requirements and add input, build, test, and repeat!

    # 3 Identify necessary data sets

    Odds are your organization, no matter the size, is collecting data surrounding most business operations. The problem is this data is often spread across a variety of different systems and software. Data may be stored in various ERP systems, MS access, SQL servers, individual inboxes, HTML pages, CRMs, databases and Excel spreadsheets. With data spread across multiple systems, getting the information you need can be an arduous task. This is where we once again need to start small and identify the necessary data sources to get started. In the end, you may take advantage of a data warehouse to improve performance. In our experience, we have seen our customers and partners start small and then aim for the larger picture.

    Remember – we will all get there but the journey starts with something small and integrating clickstream data with sales data (first small step for analytics universe!)

    # 4 Nurturing a data culture

    To get organizations on board with a new BI plan we need to evangelize the importance of data. When all departments of a company are working together towards a common goal, the resulting data insights and subsequent actions will be more meaningful and valuable.
    There may be push backs. Departments may be discouraged by a lack of time, data acumen and resources and shy away from encouraging enterprise adoption of BI or political dynamics may be a deterrent. They may not see that the adoption benefits outweigh the costs. They need to see that the right tool will benefit all teams.
    For some of the users, new technologies and platforms might be overwhelming. This may lead to change resistance but this can be overcome. Look for a platform or solution which is simpler to use with a friendly back end and with minimal requirement to code. In this age of tableaus, DOMOs and Power BI solutions, understanding insights is simpler and it’s fairly easy to access training materials.

    # 5 Finding a data champion

    Organizations across industries are moving towards data-driven culture. When we partner with our customers, we try to identify one or two data champions to work with. It helps to partner with an internal data champion as they understand the importance of driving initiatives based on data. It helps organizations and internal teams to align with the need to drive insight-driven action and not heuristics.

    # 6 – Enable Employees with Storytelling Skills

    In the end, organizations want compelling storytelling capabilities, as it is a highly effective way to help businesspeople understand data and insights. Training teams on the right skillsets and providing high quality templates are essential elements for a successful BI strategy in any organization.

    # 7 Empower Users

    Everyone within and associated with your business can benefit from contextually related data and insights. With higher adoption of insight-driven business, better business management systems are available for organizations to build on.
    Next, let’s look at some potential mistakes to avoid as we embark on the data and analytics journey.

Potential Pitfalls to Avoid

    # 1 Creating data silos and multiple versions of truth

    A few years ago, data discovery tools caught the eye of businesspeople who believed that self-service was their new information utopia. They were able to employ user-friendly tools that enabled the uploading and manipulation of personal data, followed by the visual exploration of their custom data models.

    Several challenges have emerged over time, and it usually takes two to three years of actual usage
    to realize and identify them:

    • Silos – Analytics silos are typically formed by self-service users who prepare their own data, create reports and charts, and share them within their work group. They tend to exist in a bubble, creating their own library of analytical content, with their own calculations and preferences. These silos can emerge in other parts of the organization that employ self-service tools, resulting in multiple analytical groups and multiple versions of the truth.
    • Scaling – Self-service and data discovery often start life in a department or smallish work group, where a business leader has elected to invest in a subscription for his/her team. Adding more users who need resources, security, and more data sources, can easily create a scalability problem. Self-service and data discovery tools do not inherently possess all the underlying data, content, administration, performance, user, or security management that is critical to support larger user audiences and more complex needs.

    # 2 Great-looking BI platform but low adoption leads to shutdown

    Adoption rates for BI and analytics remain alarmingly low, hovering at around 30 percent
    according to Gartner’s Cindi Howson. Many organizations have yet to reap the benefits of their
    investments because they just can’t get their users on board. We have struggled with similar situations in the past with our partners, but timeline education and constant roadshow have helped bridge that gap and improve the 30% rate to more than 60%.

    # 3 Failure to monetize data with a pay-out model

    This is a new concept in the industry and many analytics organizations don’t realize the power of their own analytics solutions. All the efforts, pain and sleepless nights can be easily monetized by building a pay-out model. Once a scalable solution is in place, use the right methodologies to charge back internal stakeholders on all analytics initiatives. Assigning a $$ value to analytics initiatives can help drive the right adoption as well as show the true value of data to larger parts of the organization.

All of us have the recipes to drive analytics success but the pointers given above are key to the success of any practice. We have seen multiple companies launch analytics initiatives but only a handful have been able to sustain for longer periods of time.

By Bhaskar Dey with contributions from Praveenkumar Sathyadev.

Bhaskar Dey
Bhaskar Dey
For over a decade, Bhaskar has worked on complex projects covering multi-channel data points in an organization's digital ecosystem. He has also been instrumental in...
Read More