CAI SIX PART WEBCAST SERIES | STEP 2
We know that data-driven organizations make much better, more consistent, and highly repeatable decisions. They respond faster to shifting customer preferences. They generate profit improvements through greater efficiencies. It’s no surprise that 98% of company leaders aspire to achieving a data-driven culture, according to New Vantage Partners.
So why are most companies still struggling to close the information loop across their organizations?
CAI, which has provided business technology services for two decades, explored this challenge in the second webinar of its six-part series: “Activating Data & Analytics for Real Business Value Creation.” The series features Steven Stone, CEO & Founder of NSU Technologies, and Tom Villani, Senior Vice President, Digital Innovation, at CAI. Stone, a former CIO at Lowe’s and Lbrands, shares real-world examples and proven solutions to help companies accelerate their own transformations.
This summary highlights key points from the second webinar, entitled: “Closing the
Information Loop Across Your Organization.” We invite you to register for the full series for free at www.cai.io and to contact Tom.Villani@cai.io for a complimentary Data & Analytics Readiness Assessment.
What Executives Need to Drive Decisions
At this point, almost all executives know they need good data to make intelligent decisions. Forrester analysts say data-driven companies are 69% more likely than their non-data-driven competitors to see year-over-year revenue growth of at least 15%. With hundreds of thousands of employees now working remotely due to the COVID pandemic, the importance of reliable and accessible data is ever-more evident.
To achieve a data-driven culture, it’s essential to close the information loop. That loop begins with the basic understanding of the information that businesses need to drive their critical business decisions and continues with the collection and curation of data to make it available for consumption.
97% of organizations feel compelled to turn their data into insights, yet 91% said that managing data is a challenge. (Gartner)
Companies often have floods of data but still lack the ability to make an informed decision. Imagine a busy business manager sorting through two dozen metrics on multiple dashboards and reports when they really need just a few data points to make a decision. Savvy decisions makers don’t waste time looking at data that doesn’t move the needle. Choose those critical metrics by talking to decision-makers, observing how they view the business, and on what data they need to make their decisions.
A good example comes from Lowe’s past when it began a major growth phase. A senior executive wanted to congratulate managers at the best-performing stores and help the stores that were struggling. Existing paper reports and dashboards actually made it harder to find that information. The simple answer was to deliver precisely the data points that identified the stores in question. That experience started Lowe’s on the path to strong analytics that drove decisions in merchandising, marketing, supply chain, human resources, and more. The chain grew to over 1,000 stores and, over the next eight years, Lowe’s revenues grew 87% while net income grew 90%.
As businesses begin, entrepreneurs often rely on gut instincts. That’s inspiring. Yet as those businesses mature, executives gather more and more data to help drive their decisions. It doesn’t mean that we completely get rid of intuition, but the reality is that gut feelings alone don’t scale very well. Culture is a top-down phenomenon. When executives embrace data analytics, it fosters a data-driven culture throughout the company.
Identify and Close Gaps in Information Collection and Curation
A museum curator has a fascinating job. The curator collects artifacts and verifies their authenticity, preserves them, documents them, and finally shares the artifacts as exhibits. Data curators perform a very similar function, particularly as the volume of data soars. We just reached 1 zettabyte of data in 2010 but hit 33 zettabytes by 2018, and we’re on track to reach 175 zettabytes by 2025. Keep in mind, one zettabyte is 1 trillion gigabytes.
Data curators need to validate their data because poor data quality can cost businesses up to 20% of their revenue. And in the age of social media, it’s also important to know what data can be used ethically. While this once represented a tedious, manual process, AI and machine learning increasingly make data curation much faster and simpler.
Once the data is curated, where does it go? The first inclination for many is to centralize it, to create a data center. But this traditional approach can be stymied by the lack of a good cleansing process and latency between the centralized data store and the operational systems. As a result, data has been getting pushed further out to the edge in federated models that support some business functions like e-commerce. Yet federation doesn’t share some information across the organization. We’re now seeing hybrid models emerge called the data hub that allows a compromise between centralization and federation.
Think carefully about choosing the right model for your organization. Make sure you know where the data is authored and governed, the latency requirements, and the need for integration. Then choose a data architecture that reflects the needs of the business architecture. A good example is Lbrands which has five different brands including Victoria’s Secret, Pink, and Bath and Body Works. It developed a master data hub strategy before that term was popular, using APIs to connect business processes through a mediation layer. That model included a “P hub” for products and a “C hub” for customers, even though the different brands provided very different products to different types of customers. As a result, data quality increased dramatically and the hubs greatly simplified the time needed to drive projects thanks to a high degree of reuse of product services and APIs.
Remember that each business user absorbs information differently. A highly visual person wants graphics with the ability to drill down, but another person may want a high-level grid report that delivers text alerts. Data must be curated, stored, and delivered in a unique way for each business to truly support the best business decisions.
Incorporate Analytics Into Existing Processes
Even when you make reliable data available in the most appropriate form to the decision-maker, there’s a question of how that analytics capability can be melded into long-existing decision-making processes. The simplest way is through decision support – the entry point to analytics.
One proven approach is to build on what’s already there. If a decision-maker has relied on a trusted report for many years, simply add new features. Be sure to eliminate the noise, as discussed earlier, to deliver the most critical data points as close to the decision-maker and the decision as possible.
By 2022, more than half of major new business systems will incorporate continuous intelligence that uses real-time context data to improve decisions. (Gartner)
The next approach to consider is decision automation. Business intelligence and decision support aren’t enough when decisions need to be made in real-time. With the rapid emergence of AI and machine learning, the ability to automate decisions is growing exponentially.
Not only do many software solutions now build automation into their products, but the big three cloud computing providers – AWS, Azure, and Google – have made decision automation much simpler and more powerful.
The problem is that most decision-makers aren’t ready to trust a Black Box to make a big call. Instead, business managers are warming up to intelligent advisors that support complex decisions by analyzing a wide range of variables, data, and trends, then running various scenarios before proposing a solution.
Data-driven organizations are seeing 20%-30% improvements in EBITDA due to unlocked efficiencies and more granular financial insight. (Forbes)
Lbrands took this approach, even incorporating a voice interface to communicate with the decision-maker. For example, it might make a recommendation for distributing remaining inventory on a Monday morning based on the weekend sales trend. It literally tells the manager of its ideas. Then the manager can run additional analysis using other data points. Finally, the manager adopts and implements a solution. At the end of the week, the intelligent advisor can compare how well the two models worked.
The result is akin to a closed-loop audit on the decisions that were suggested and those that were made. This also has the advantage of helping a new manager quickly start making informed decisions based on the years of experience of the company’s top merchants.
The concept of a closed-loop audit also applies to those developing analytical solutions for business users. Technology project managers traditionally complete one project, then move on to the next one on the priority list. However, as business conditions shift, business managers using analytical solutions tend to augment the system with data from other sources. Or they may stop using a solution entirely. To ensure the loop is closed, technologists need to audit usage. Sometimes, that requires them to take a step back and consider what else might be important.
Project managers often use colors to monitor progress in product development. All too often, a project jumps from “green” to “red” and hits a hard stop. That suggests the developers missed something along the way. To combat that, it’s good to use a system that employs an intelligent advisor that serves as an early warning system.
A real-world example is CAI’s home-grown solution, TrueProject®, which collects hard data from project management tools and combines it with qualitative data collected through automated project assessment capabilities, based on the input of key stakeholders. That input is collected in three- to five-minute surveys of all stakeholders and blended with hard project data on a continual basis.
Analytics built into TrueProject determine conformance with best practices, pinpointing areas where the project may go off track. An AI component develops recommendations to help the decision-makers take preemptive action. The system uses a rich set of visualizations to help decision-makers respond to the situation, making it a perfect example of a closed-loop of information delivery.
The point here is that deploying a closed-loop solution requires the support of business stakeholders and analytics specialists to build iterative models that can evolve over time. By closing the loop, they ensure the analytics process they build remains relevant to decision-makers who rely on them.
We'd love to talk.Get in Touch