Create value from data by taking a results-obsessed approach to analytics

Executive Summary

• For executives, there is a large gap to close between investment in data analytics and ROI 
• The secret to actionable results is using data to solve business problems—not building massive analytics capabilities because the technology is possible
• Investments will continue to fail if the human aspect isn’t addressed—business leaders must get more comfortable with analyzing, digesting, and using data in their everyday work
• In our experience, two specific and defined approaches have proven to jumpstart analytics capabilities that drive results—regardless of company size 

Introduction

When the economic crisis from the COVID-19 pandemic came on, companies scrambled to make sense of the situation. And as reality set in, many leaders struggled to mine—or even access—their data for critical decisions. What would happen to cash flow if clients stopped paying their bills? How do they staff if demand spikes or plummet? What impact would restrictions have on their supply chain for one month, or six months? 

The situation, for its wide scale and acute importance, exposed a serious issue: Too many companies have built analytics capabilities based on technology possibilities—not to solve real business problems.   

The fact is, any leader, in any industry, at any size enterprise, can learn from and act on data. Not through massive upfront investments in new technology or platforms, or system overhauls that devour millions of dollars, but by rapidly extracting data to answer questions and provide insights that matter to their businesses today. 

Everyone is hyped up about machine learning and artificial intelligence, but that excitement distracts from the urgent need to get value out of data. Plugging in any technology without a specific purpose will not magically create value. And there is a lot at stake: Companies are spending $189 billion on big data and analytics, according to IDC, and that figure will grow double-digits for at least the next two years. 

The goal is to create value with data by extracting insights that empower seasoned operators to make smarter decisions. When those decisions turn into actions that drive down costs, make teams more effective, or lure and keep more customers, value gets created—and the business get transformed.  

Chapter 1: The gulf between investment in data and actionable results

The last several years have produced a regular drumbeat of headlines from publications like The Economist declaring that “the world’s most valuable resource is no longer oil, it’s data.” 

That infamous cover story set off a frenzied wave of investment across the business world. But the urgency to act, mainly by investing in data collection and infrastructure, has done more harm than good, thanks to erratic data strategies that have swallowed millions of dollars while producing little value for the businesses. 

That helps explain why surveys of executives revealed a persistent gap between those who want to be “data-driven” and those who actually are. The problem is those leaders are fixated on building massive data infrastructure when they should be starting small by seeking insights that cut straight to the bottom line right now. That’s how they can close the gap between investments in data and the results they generate.

A recent study showed that 92% of Fortune 1000 companies are increasing big data and AI-related investments, and that the same percentage of respondents say the investments are required to remain agile and competitive. That means fear of competition, rather than specific business-use cases, is motivating their investments. But only 62% of enterprise players report generating measurable results from those investments. 

While this study is useful in illuminating the gap between data investments and their results, it stops short of showing how successful data analytics initiatives can deliver competitive advantage. They can, but leaders first must understand how to close the gap between investment and results.

It starts with asking questions that align to strategic goals, but that will produce answers you can turn into tactics. If the goal is improving client retention, questions might seek to learn at what exact points in the client journey they fall away or opt out. Data analytics can reveal the answers, pointing to upgrades and fixes that will quickly improve client retention. 

The truth is that no matter how advanced your tools and technology are, your data will not deliver value unless you extract insights that drive strategic results. And it all starts with asking impactful questions about your business: How long does it take prospects to become customers? Are customer service interactions improving or decreasing customer satisfaction? Why do customers churn? When do they churn? 

Once you have baseline answers, you can craft hypotheses related to the business. Looking at the data will help you prove or disprove them, enabling your team to tweak variables and measure results. This is where data is most useful. In the beginning, those hypotheses can and likely should be small in scale. If we do X, we think Y will happen. And we are going to measure Z to tell us if we were right. 

Chapter 2: People + Process > Technology

We advise our clients to stay true to their DNA. If you’re not a technology company, don’t try to be a technology company. Avoid the technology arms race, and use your precious time, capital, and talent on data analytics projects that move the needle for your business. 

Just as the majority of executives say their investments in big data and AI aren’t producing results, the portion who consider their organizations data-driven has also declined, from 37.1% in 2017 to 32.4% in 2018 to 31% today. The same survey reported that only 7.5% of companies face technology-related obstacles to implementing big data and AI initiatives. Instead, cultural issues—people and processes—represent the bulk of the challenge today. 

Successful data operations that yield actionable business insights are less about specific technologies than about the people. Simply put, business leaders need to get more comfortable integrating data into decision-making— seeking out the right data, working with data analysts, and using critical reasoning. In short, data needs to be driven by business—and more specifically, businesspeople—but enabled by technology. 

If the information being collected and analyzed isn’t directed at a specific business problem, or employees are not consulted and trained on data analytics, or leadership teams are not in place to act on insights, investments fail to deliver. 

A major utility learned this the hard way after spending tens of millions of dollars on a proprietary analytics platform, filling up servers with every byte of operational data they could get their hands on. But just a few years in, their massive upfront investment is simply not paying off—the collected data has been languishing on those servers. Their error was diving straight into custom, proprietary technology without thinking through organizational enablement.

Companies invested $1.3 trillion on digital technologies last year but an estimated $900 billion of that failed to deliver the intended results—because 70% of largescale, complex change programs fail. That’s because most digital technologies only provide possibilities for efficiency gains and better business outcomes. But if the organization—and its people—don’t adapt and its processes are flawed, progress will be stifled.

Too many executives still believe judgment, experience, and gut instinct are the key factors in good decision-making. We get it—we face the same challenge. But we all should become more curious and ask questions instead of relying on traditional assumptions. Data leadership is crucial to creating value—and it should come from the top. 

Data must be driven by the business and enabled by technology. The most enlightened data organizations exhibit this, with C-suite executives and directors directly setting the data agenda according to their most pressing business challenges and priorities.

We recently helped a large global insurance brokerage revitalize their data analytics efforts by working with them to identify and evaluate 16 AI use cases. Rather than reaching for the most impressive-sounding use case or seeking the most advanced technology, we worked alongside data engineers, business leaders, and sales leaders to identify a particular data-enabled use case that would be simple and quick to achieve . We spent critical upfront time exploring the data feasibility, playing out models, and learning about the real-world impacts from on-the-ground operators. The result: a projected $35 million boost to revenue. 

Organizations that keep trying to introduce new tools and software only to produce middling results should also be concerned about employee frustration, reduced productivity, and wasted capital. In a 2019 West Monroe study, “The Upskilling Crisis,”  we found that 84% of organizations sometimes or never redesign an employee’s process before introducing new technology to their role. We often engage with companies that are spending millions on new systems but then skimping on training and support for employees.

Here’s what a focus on process redesign could look like: Develop a use case envisioning what the day-to-day would look like for a new data collection and analysis process. Which departments and what roles within those departments will be involved? Having identified those people, loop them in to understand what specific process changes would need to happen for them—and what obstacles they would face—to realize value. Without this level of “homework” and understanding before redesigning a process—and certainly before applying a technology or tool—the full value of data investments can’t be realized. 

To illustrate this, here’s one example we experienced with a client: After leadership implements predictive modeling software in a call center to suggest customized upsell offers to operators, call center representatives could decide their own experience and judgment works better than the automated prompts, and simply ignore the suggestions. Without training on the technology and trust in the output, the investment in the predictive modeling will not yield its desired results. 

Chapter 3: Asking the right questions and homing in on hypotheses

Asking the first few questions is the hardest part. Answering the questions with data will generate even more questions—one question begets five more. This creates a beneficial flywheel effect but requires discipline to track questions and systematically use the right tools to get the right answers. 

Data analytics should adhere to the rules of supply and demand. IT professionals and consultants want to supply data, but if others in the organization don’t know how to use it, there won’t be any demand, the data will go unused, and producing it won’t be worth the investment. 

Start with demand instead. What kind of information do people need? Persuading internal stakeholders and continually sharing success stories will stoke demand for data and prompt employees to flex their curiosity. Start them thinking about what kinds of information would help them make better decisions daily. The key here is developing hypotheses that can yield rapid insights to see and demonstrate benefit within four to six weeks. 

Start with a burning issue—Why do employees leave? Why do upsell attempts fail?—and create a hypothesis. Then create another until you have four or five hypotheses you want to test. If we adjust price by 5%, will that increase sales 2%? If we email new prospects once every three weeks instead of six weeks, will conversion improve? Start with specific queries and put a stake in the ground, and keep testing until the data gives you an answer. 

Testing multiple hypotheses at the same time is worth the additional complexity; it provides backups, scale, and will build your capacity and proficiency at managing multiple tests at once—a key benchmark of a healthy analytics organization. 

To coax value from data (and not lose your patience), direct inquiries at business issues. When you start pursuing efforts to seriously work with data, the questions should be narrow and discrete. Start humble with simple questions. Answering those questions with data over time will build up confidence in bigger investments and efforts.

When large investments are made without the people, processes, and cultural infrastructure to guide how you’re going to use the data, that capital is destined to be wasted. Starting too big causes confusion and paralysis. 

Chapter 4: Moving strategic levers and building rapid insights

When we work with clients, we typically start with one of two approaches, depending on the size and maturity of the organization: 

Approach 1: Moving strategic levers

The strategic approach starts with aligning data initiatives to an organization’s strategic pillars. We then help them create a timeline of initiatives, mapping each initiative to one of three buckets: people, processes, or technology. 

• People - Do your people have the right mindset and skills to interpret data results? 
• Process - Is the data you’re collecting and analyzing solving a specific business issue? 
• Technology - Do you have the right tools and software in place to collect the data you need? 

Based on this exercise, on-the-ground stakeholder input, and available technology, we make recommendations on how to realize the most value from a data initiative. The important thing is, we start at the beginning with a large, strategic picture of analytics—not a proof of concept. 

Here’s how we demystify data analytics with this approach: 

• Workshop and prioritize key questions with top-level business and tech leadership 
• Execute a rapid analytics cycle (2-3 weeks) to get answers depending on priority of questions 
• Maintain cadence of original stakeholders, update team with insights and overlay potential business levers that could get value from the insight 
• Manage strategic vs. tactical insights to inform quick wins vs. large capital or strategic moves 

Approach 2: Rapid Insights

Rapid Insights is a different technique that can be used to jump-start analytics. It is simple and tactical—and it is not strategic like the first approach. This program is designed to provide an on-ramp between an organization’s tech and business teams and our data experts.

It starts by producing preliminary data visualizations within six to eight weeks. This enables business leaders to quickly see a rough draft of the fresh insights available from their data and allows them begin making behavioral changes. Alternatively, if the analytics do not match the organization’s needs, Rapid Insights allows for data architecture modifications—hypotheses are verified before significant data modeling investments are made. 

These analytics sprints are refined and repeated, delivering insights every step of the way until the analytics and data governance are optimized. But the process involves much more than static reporting. Just as data is compiled in nimble sprints, management information and data science are applied to quick behavioral experiments to see if they can produce value before they are operationalized. Because the process is incremental and continuously delivers results, organizations can quickly see value, generating momentum for the initiative over the long term. 

Often, we help companies start with Rapid Insights to see and experience the value of data. We use that experience to go back to the strategic approach that sets up a long-term roadmap of data initiatives.

Conclusion: Starting your analytics journey is not as hard as you think

Data analytics is a huge, overwhelming topic that leave seasoned executives paralyzed with indecision. Instead of jumping into the deep end of the data pool, step into the shallow end. Sometimes it’s better to define a larger strategy that sets the stage for all future investments. Other times it’s better to take a rapid analytics approach to get insights fast, prove the value, and take a step back. We have seen both approaches work for clients. 

Whichever approach works for you, take steps to advance data operations in your organization that focus on culture, people, and processes. Here’s how:

1. Build a data analytics culture

Starting with narrow, specific questions helps people get comfortable with the process of making evidence-based decisions from data. Start asking executives what they wished they knew. Put simply, “What do you wish you knew about your business, your customers, and your employees?”

2. Turn to the people you already have

Think through a skills and gaps assessment. How do you get from the workforce you have to the one that you need? How can you utilize effective organizational change management to promote continual and consistent learning to meet digital transformation goals?

First, clearly define a vision for what you want to achieve in the near -term and articulate the business value of that vision. If achieved, where will the business be? Be specific with the metrics.

In order to attain the future state, what will need to change in the business? How will the organizational structure shift? Consider what happens to the business and operating model. 

Finally, think through what skills are needed to achieve the future state. If that future state is more dependent on data and analytics, that will likely be an area where you need to increase workforce capacity.

3. Reimagine processes to operationalize data analytics

Once you’ve instituted a culture of asking data-driven questions and have the right people with the right skills in place to analyze answers, process is what brings it all together and enables your operations to scale. 

Make sure that data leadership and ownership start and reside in senior business leadership. Work toward creating an internal Center of Excellence, a cross-disciplinary team of business analysts, operations, data scientists, and IT leaders. Together, this will be the team that drives results.