Home IT management Building A Data-Driven Organization

Building A Data-Driven Organization

0
Building A Data-Driven Organization

[ad_1]

Building data-driven businesses is clearly the topic of most board agendas this year. Increasing technology maturity coupled with the macro-economic environment is driving a sharp pivot to accelerate data monetization and define analytics as a product for many organizations. The reality is that creating a data-driven organization is easier said than done. Research indicates that just 23 percent of executives reported that their companies had created a data-driven organization, down from 31 percent four years ago.

Here are three areas to consider on the journey to building a data-driven business:

Create an enterprise data platform

Building an enterprise data platform (EDP) with “truth” in data is key to unlocking the value of data across an enterprise. The before and after pictures of companies that started with fragmented data with ownership across distributed groups and moved to a single centralized governed EDP are strikingly clear. The clear and winning model that has evolved is “data at the center, analytics at the edge.” There is evidence that cloud is actually exploding the total cost of ownership when it comes to data – and volume is the single largest problem many in the industry are facing. Breaking down the volume problem by focusing on quality over quantity is key. Leading corporations are setting up data intelligence factories – a single place to manage, certify, and publish their data globally. Once in place it can be coupled with a front-end interface for business leaders where they access, and export trusted and reliable business data, directly from the source.

Governance is key in this matter – being intentional about the way data is managed, governed, controlled, used, and owned. We’re often so focused on the outcome that we forget that the input is really critical. Making data valuable requires the right governance and ownership – data strategy, lineage, and clear lines of ownership – with accountability flowing back to the business. In addition, for data to deliver truly meaningful output, data needs to be contextualized at the industry or subindustry level, and building in a foundation for this context is crucial for today’s data-driven enterprises.

But even for companies that are truly data-driven, the day-to-day management and governance over data quality and engineering is still an evolutionary discipline. In many ways, this is a journey, not a destination, and there needs to be a fundamental and continued willingness to learn, experiment and innovate, to get to a true data-driven business status.

Develop a clear strategy for data science teams

With the proliferation of data, there is a need to prioritize efforts of data science teams along a portfolio approach, keeping a strong focus on key stakeholders and learning when to follow, partner or lead, as discussed in a recent interview with Murli Buluswar, head of analytics for Citi.

The portfolio approach is tried and tested in the venture capital world, but it applies to data science as well. The data science organization needs to build relevance in the short, medium and long term. You have to be tackling a set of problems that satisfy the business in the near term while ensuring that you’re driving step change intelligence in the medium term and then more fundamentally, transformation in the long term. Having that portfolio approach allows you to be strategic and relevant. That is true for capital allocation as well.

Putting yourself in the shoes of your key stakeholders is also a key success factor for a data-driven strategy. If you are thinking like a CEO, you’re thinking of materiality. How is this bending the curve on the future of that business unit or the larger enterprise in a way that is meaningful and appreciable at scale? If you are thinking like a CFO, the measurement manifests in either the P&L or the balance sheet. And if you are thinking like a Head of Audit, you have the mindset of assessing whether the decision is having the impact that it intended to. Some examples of outcome metrics include 1. financials realized 2. financials identified, but not yet realized. 3. adoption and non-financial change, for instance the speed of decision-making in some particular area, or more clarity with which decisions are made. 4. new frontiers of innovation, new questions that we are asking that are more early stage and will hopefully manifest themselves in outcomes and financial metrics. 5. the beta of what we do, delivering on your operating commitments a critical part of ensuring that the entire ecosystem is operating effectively.

Looking at analytics and data science end to end is key. The risk for many data science functions is to measure success through a simple functional lens of delivering a set of insights. But those measures are an intermediate step, they are not necessarily the end metrics. A CEO might care about the fact that his or her decision sciences or analytics team is coming out with super useful insights, they are more focused on this is driving financial outcomes, the breadth or the speed or the depth of decision-making in some important area of their business. That requires the data science capability to not just think vertically as a function but think horizontally and understand the end-to-end process.

Establish a data-driven culture

The need to create the right culture within the business is key – and the right culture is one that understands the value of data. Data is only valuable if you do something with it. And the best technology leaders play a large role in helping the business really understand how they can use data to achieve better outcomes, asking the questions the business does not know to ask. There is a need to move from data science to decision science within the enterprise. With this new focus, the organization is not just chartered with delivering analytics or insight, it is chartered with delivering clarity on the decisions being made. It therefore needs to have an acute sense of the outcome you’re trying to drive.

One great example value of data-driven superior outcomes is digital twins – instead of bringing a manufacturing line down and re-laying the parts and pieces, the ability to assemble, run and test a digital twin, before even touching the assembly line, presents significant cost savings for businesses and reimagines how enhancements are delivered.

The other critical success factor is culling ideas quickly and decisively and only developing and driving initiatives that will have an impact at scale and within a reasonable timeframe. Along with the data democratization we want to see, new issues emerge, especially around curiosity, accountability, ownership, and change management – often clubbed together as “data literacy.” A framework can accelerate data literacy programs across the corporation, and the first piece is new tools that are designed for businesspeople instead of just data scientists. The second is driving agile programs across the company that demonstrate the journey from ideation to visualization to outcomes. The third is affecting the mindset of making data a first-class citizen. And the best practice here is driving mindsets top-down, not bottom-up, starting with the CEO’s office.

[ad_2]

Source link

LEAVE A REPLY

Please enter your comment!
Please enter your name here