5 'No-Regret' Moves for becoming Data-Driven

No-regret moves are actions that pay off no matter what happens, and are useful when deciding on a strategy under conditions of uncertainty. In the context of becoming data driven, these are initiatives that require relatively smaller investment, but with potential larger benefits in terms of generating business value, improving operational efficiencies & reducing risk.

Becoming data driven can be challenging as evidenced through a survey of select Fortune 1000 industry-leading firm C-suite executives, conducted by the Harvard Business Review in 2021. ?In the survey, companies reported a marginal decline in the leading metrics that are used to measure success of their data & AI investments. Companies reported struggling to make progress, and in some cases - losing ground on managing data as a business asset, forging a data culture, and using data to drive innovation.

The journey to become data driven may seem daunting depending on current state of organizational data maturity. It is prudent to embark with certain no-regret moves that require relatively moderate levels of capital outlay.

Following are 5 no-regret moves for becoming data driven:

1. Data & Analytics Centre of Excellence (CoE)

It is important to inculcate culture that leverages data for all important actions, communications & decision making, right from the top of organization. The senior leadership sets the tone on essentiality of data in operations & decision making, that then propagates through the organization.

?The CoE facilitates this by promoting data driven culture that includes selling the business value through driving organizational awareness & ideation. Value propositions are developed & communicated across the organization. The current state of data literacy & cultural readiness is assessed, and suitable interventions in the form of learning & training plans are devised to address gaps vis-à-vis the envisioned target state.

?The CoE should promote creating cross-functional teams involving functional roles & data scientists, for example, by having data scientists take up line roles for specific projects.

2. Self-service Tools for Analytics & Business Intelligence (BI)

Gartner defines self-service BI as end users designing and deploying their own reports and analyses within an approved and supported architecture and tools portfolio.

?It allows business users to explore, access, understand, and utilize data effectively. It is enabled through a business-friendly & governed semantic layer for holistic & consistent view of data and leads to greater innovation & collaboration across business communities through sharing of concepts & ideas.

Self-service analytics & BI tools have smaller resource footprints & require nominal IT support, while still providing powerful feature sets & allowing for adequate governance & controls.

3. Data Science & Analytics Sandbox

A data science & analytics sandbox facilitates experimentation & provides flexibility for leveraging tools & technologies that may or may not be part of certified corporate technology stack. Cross functional teams of experts can collaborate within the sandbox environment to bring new ideas to life & new data-driven approaches to complex business problems.

?Sandboxes should be free of standard constraints such as data modelling standards etc. to facilitate exploration of data in an environment that doesn’t impact daily operations or tie them to existing business infrastructure. Innovation teams need to understand, embrace, and apply the concepts quickly to explore new business solutions and products that will differentiate their offerings.

?Data scientists & citizen data scientists are enabled to source, integrate & aggregate data from internal as well as external data sources to build prototype algorithms, data solutions & advanced analytical & predictive models. The models developed can be implemented to production environments for enterprise-wide adoption, with faster time to market & minimal upfront costs.f

4. Proof of Concept (PoC) for ‘Fail-Fast’ approach

Data-driven PoCs powered by realistic & contextual data should be adopted for developing business cases & validation of assumptions & business ideas. The PoCs should be robust and scalable for production readiness.

?The initial business use cases should be selected after consideration of business outcome & feasibility of timebound implementation (ideally within a duration of 4 to 6 weeks). Expectations should be detailed at the outset, through clear metrics for success by listing qualitative and quantitative measures.

?5. Hackathons

Data-focused hackathons can be leveraged for innovation & generation of ideas, with the goal of creating a functional data product at the end of the event. Teams collaborate to develop a proposal, build prototypes & pitch ideas to management. Shortlisted ideas & products are taken further for implementation, based on feasibility & potential for value generation. Hackathons facilitate multi-disciplinary collaboration, employee engagement, learning, that goes towards promoting data-driven culture.

Conclusion

Thus, organizations can start the journey to become data-driven by focusing their efforts on high impact ‘quick wins’ through these no-regret moves. Becoming data-driven is a long-term process that requires continuous & persistent efforts, and the payoff in terms of generating value out of data can be significant.

?