Five Analytics Challenges Most Organizations Face and How to Overcome Them

Think fast. You have an opportunity to buy a portfolio that could help your organization successfully enter a new market. Expanding into this market will give you an important strategic advantage over your competition and potentially open doors to other expansion opportunities. Upon preliminary review, it’s unclear whether the portfolio is in line with your current risk strategy and managing it might overwhelm your systems. This could both jeopardize your opportunity and add risk to your existing portfolio. It’s a big bet and you have a limited amount of time before the offer expires and the portfolio is offered to your competitor. How do you decide what to do?

There’s more than one way to approach this opportunity — and going with your gut isn’t one of them. Given the magnitude and complexity of the decision at hand, advanced data analytics must factor into your decision-making. But do you have the necessary organizational analytics foundation to make an informed decision? Have your key stakeholders fully bought in? Do you have the right team? Are you able to effectively access, manage, analyze, govern, and put your data to use? Understanding these common challenges and taking a pragmatic approach to overcome them can set your organization up for long-term success.

Today’s challenges are yesterday’s challenges – but bigger

The proliferation and accelerated adoption of digital commerce and communications has triggered a data explosion. It's estimated that 2.5 quintillion bytes of data are created every day and businesses collect great volumes and varieties of data with every transaction, interaction, website visit, and social media reference. This data holds vitally important information to help financial institutions, for example, guide decision-making, fraud detection, marketing strategy, etc. Yet despite the widespread advances in technology and techniques, data teams continue to struggle with even the most basic functions, including wrangling, cleaning, and preparing the data for analysis. These and other challenges are major obstacles to harvesting, let alone using, the critical insights within the data. While seemingly obvious, it’s important to understand these issues and why they’re persistent so you can figure out how to work with — or around — them.

1.   Lack of organizational support

The biggest and most important challenge to overcome is organizational buy-in, typically in the form of resource allocation. Many businesses continue to be driven by legacy visions and values, intuitive decision-making, or both. They have not (yet) embraced data analytics as a strategic priority and typically view data management as a production group, under the oversight of IT.

How to approach: Start with the end in mind

While your organization may face some, all or none of these challenges, the return on your analytics investment (e.g., data, technology, people) will be hard to calculate, or even envision, without a clear articulation of what you aim to achieve. Starting with the end in mind provides the basis for a strategy for how to get there by knowing where to focus and what to fix. It helps you articulate what’s important, what you need, and sets the stage for gaining the support and resources to achieve your goal. What stands in your way?

Make a business case by helping stakeholders better understand the value advanced analytics can deliver. Build your approach around use cases aligned with business priorities and strategy to win the attention and interest of the leaders responsible for delivering business results. Here are a few examples where analytics can be of use: 

• Market intelligence and research: Identify areas for growth and build strategies to capitalize on them
  • Define consumer segments, regions, and products to inform strategy
  • Size segments/opportunities to prioritize investments
  • Deliver unique benchmarking insights based on a “full wallet” view of consumers
• Customer acquisition and engagement: Create positive customer experiences to foster new and nurture existing relationships
  • Supply a holistic view of the customer to help build personalized experiences
  • Identify profit and loss drivers, channel profitability and cross-selling opportunities to inform marketing strategy
  • Refine, validate, or build new, custom underwriting models
• Account management: Enable preventative intervention to manage potential delinquencies
  • Analyze how specific vintages are performing versus competitors to guide management and treatment strategies
  • Identify pockets where credit quality may be deteriorating
  • Optimize prioritization and contactability models to improve collections
• Risk management and fraud protection: Securely manage risk and regulatory compliance while promoting transparency
  • Supply a robust view of business (past, present, and future) to better identify and manage risk
  • Find and rate individual customers at risk of fraud and apply different levels of monitoring and verification accordingly 
  • Inform how to incorporate technology into compliance and risk management processes

2.   Insufficient data science skills

To derive the insights necessary for decision-making, organizations need to have the right mix of data science, analytics, and business intelligence experts with a balance of roles and skills. Market demand for competent data professionals was predicted as early as 2012 when venture capitalists (VC) were worried enough about the tight data labor pool, they built specialized recruiting teams to feed their businesses. Supply continues to outpace demand with businesses of all sizes facing pressure to attract and retain this talent. According to one survey, 36% of respondents cited recruiting and retaining data scientists as the biggest challenge to achieving their company’s AI goals, besting both inadequate technology (28%) and insufficient data for model training and accuracy (26%).

How to approach: Bridge the gaps

Upskilling current employees and hiring to fill gaps is an important component of building a data-driven culture and organization but it is a process that requires commitment, discipline, time, and money. Alternative ways to reinforce your analytics capabilities and contributions include process automation and outsourcing expertise. Advanced analytics consultants, for example, can be an expedient and efficient way to supplement and fill knowledge and skills gaps.

3.   Accessing, connecting and securing data

Like the links of a chain, data-driven decisions are only as strong as the weakest datasets used to make those decisions. Data integrity — the property that data has not been altered in an unauthorized manner — is essential for reliability. If the data is corrupted, modified, or destroyed, integrity is compromised and trust in the data will be lost.[5] The downstream repercussions of decisions made based on the compromised data could be extensive and enduring.

The chart below is more than seven years old but still representative of how data scientists spend their time. Numerous articles and studies continue to report data scientists spend 60% to 80% of their time collecting and cleaning data rather than analyzing them.