In recent years, unprecedented disruption from the COVID-19 pandemic and geopolitical tension has forced businesses to rapidly evolve both their management processes and their business models.
In today’s fast-moving and uncertain environment, the enabling general and administrative (G&A) functions of a business—such as HR, IT, and finance—increasingly need to provide rapid insights into ever larger and more complex sets of data to guide decision making and drive business performance. Without a firm grasp of advanced analytics, companies could struggle to understand how their business fits within the broader market and what can be done to ensure they remain competitive and successful.
Yet G&A functions face multiple challenges in implementing advanced analytics. Because companies traditionally view G&A functions as a cost center, finding the money advanced analytics requires is an uphill battle—particularly because it is difficult to show a direct correlation to top- or bottom-line growth. While there is often an indirect impact—for example, better forecasting can result in a more effective allocation and deployment of business resources—translating these effects into a compelling business case can consume resources that functional leaders would prefer to use in funding and scaling new analytical capabilities.
A new McKinsey survey of more than 300 leaders of various corporate functions (including CFOs, chief human resources officers, chief information officers, and general counsels) provides fresh insight into which types of analytics are being used by different G&A functions—and, even more importantly, which factors increase the chances of successfully introducing advanced analytics to provide business with powerful information on how to perform better.
The rise of advanced analytics in G&A
The survey indicates that over half of corporate leaders (53 percent) are increasing their advanced-analytics investments for G&A functions, while only 1 percent are actively cutting back investments compared with the previous year.
These investments appear to be paying dividends, resulting in a significant uptick in the usage of analytical models and techniques within each function. This is happening most markedly in the functions that were most challenged operationally during the pandemic, especially because of increased labor turnover and the sudden transition to remote working.
For example, leaders of functions such as real estate and HR indicate that, over the next 12 months, they expect to almost double the number of different applications for analytical techniques, by 80 percent and 73 percent respectively. Even functions that have long relied more heavily on analytics—such as procurement and finance—are applying analytics to a broader set of situations, with usage expected to increase by 40 percent and 21 percent respectively.
A significant percentage of corporate-function leaders also report using more sophisticated forms of analytical techniques, beyond basic reporting, to provide usable insights to the businesses they support (Exhibit 1). The procurement function seems to be the most advanced in this regard, with more than 40 percent of chief procurement officers reporting they are now performing some form of predictive or prescriptive analytics to guide decision making.
Other functions are rapidly catching up. For example, real-estate functions tasked to support a rapid shift to hybrid work are increasingly analyzing building attendance and remote-work data to shape their future building portfolios. In tandem, HR is bringing analytics to bear on issues such as workforce planning, candidate screening, and talent attraction (Exhibit 2).
These investments can be transformative for businesses (see sidebar, “Advanced analytics in action”). Since the mid-1990s, McKinsey research has quantified a company’s “analytic quotient”—a measurement of the maturity of the company’s analytics deployment, the effectiveness of its technology and analytics models, the strength of its data management, and the depth of the analytical skills among its employees. At companies with a higher analytic quotient, revenue growth is more than double that of peers with typical analytical capabilities, and five-year total shareholder returns is two and a half times higher.1
The challenge, nevertheless, is daunting, especially for businesses starting from scratch. However, the survey reveals two ways for a company to begin its advanced-analytics journey.
Identifying and selecting potential analytics applications
Companies face a complex decision-making process when deciding which functions and use cases to prioritize for advanced-analytics investment. Assessing the organization’s maturity in developing and deploying analytics can help pinpoint the best strategy and place to begin. Depending on the results, there are two routes a company can take to implement advanced analytics: adopting tried-and-tested analytics techniques or building a bespoke model.
For less mature organizations, research suggests a test-and-learn approach—one that allows staff to roll up their sleeves and start experimenting with potential use cases—may be an effective way to set up analytics capabilities by adapting examples that have been widely implemented elsewhere. For example, an HR department may want to start by focusing on the function’s most basic analytics applications, such as understanding drivers of retention, predicting the future workforce’s skills needs, or establishing predictors of job performance. With experience, more complex analyses can be added in subsequent waves. At the intermediate level, talent sourcing, salesforce effectiveness, management effectiveness, and employee motivation are potentially worthwhile analytics targets. The most complex and ambitious HR use cases include predictive hiring and screening, absenteeism forecasting, succession planning, and time-and-expense auditing.
Would you like to learn more about our Operations Practice?
While starting an analytics journey along the best-trodden pathways may be the most expedient option for many organizations, companies that are more confident in their analytical abilities may be able to realize value faster through a more formal prioritization process. This approach, which requires more time and resources but can be tailored more specifically to the company, applies a set of four filters:
- How well could the analysis answer the most important questions inherent in meeting the business’s strategic objectives?
- What business impact could result—such as increased revenue, better-managed risk, or reduced operating cost—from the decisions that this analysis would inform?
- How vulnerable would implementation of the findings be to potential barriers—such as unavailable or poor-quality data, external dependencies, or legal considerations (such as data privacy)?
- How much will it cost—including for system infrastructure, analytic and visual tools, and human capabilities?
By following this approach, stakeholders at a global agrochemical company were able to identify the people- and financial-management issues where analytics would matter most to a solution. For HR, improving performance management and staff retention were the top objectives, while finance focused on better demand forecasting, improved payables performance, and more accurate cash forecasts. Channeling resources to these priorities increased the odds that the analytics investments would pay off, building confidence and capabilities.
Starting and scaling advanced analytics
Our survey of CXOs across a variety of industries, functions, and geographies identified six broad categories of enablers at the organizations making the best use of analytics in corporate functions (Exhibit 3).
Some factors had a greater impact than others.
Organization. Among surveyed companies, setting up a dedicated analytics center of excellence (COE) was the single highest-impact factor on the deployment of advanced analytics: organizations with COEs applied advanced analytics to 55 percent more use cases than those without. This pattern held both where the COE focused on a single function as well as where the COE worked across multiple functions.
Data management. Not surprisingly, ensuring data quality also proved important. Organizations with strong data-governance and data-access practices were able to deploy analytics 43 percent and 40 percent more frequently, respectively, than organizations that rated themselves less capable in these disciplines.
Technology. Analyses aren’t worth much if decision makers can’t understand them. Investments in data visualization and communication technologies, including in self-service platforms, can therefore make a critical difference. But technologies for achieving perfect data quality proved less decisive, suggesting that budgets could be reallocated elsewhere.
Analytics models and tools. Organizations that described their data-modeling platforms as “robust” delivered 16 percent more analytics use cases than their peers. Departments with tighter budgets, or at the early stages of the analytics journey, could consider prioritizing technologies that improve access to—or better visualize—data before allocating significant resources to building sophisticated analytics platforms. These tools are perhaps better reserved for organizations that have already benefited from the more foundational technology investments.
Strategic alignment. Leadership alignment on the value of deploying advanced analytics, together with a clear funding mechanism for necessary technology and talent investments, helps support wide adoption of analytics. But survey findings also uncovered a law of diminishing returns. Focusing too much on tactical prioritization, at the level of specific use cases, actually inhibited the broad use of analytics within an organization (Exhibit 4). Especially in the early stages of building and scaling analytics, the right balance appears to be a test-and-learn approach—rather than wasting time trying to align all stakeholders on which use case gets implemented first.
The right kind of talent. There are three major skill sets to consider for advanced analytics: expertise in analytical techniques, the ability to translate business issues into questions that can be answered with analysis, and data engineering capabilities. However, each had only a relatively modest impact on the rate at which corporate functions scaled analytics. The implication: a lack of depth in these skills is not the impediment leaders might think—making the test-and-learn approach practical for more companies.
Getting started and building momentum
For companies starting on their analytics journeys, there is encouraging news. While the six governance practices all contribute to success in standing up and scaling analytics in functional organizations, progress in even one or two can help build crucial confidence. Early on, the focus might be prioritizing the levers with the greatest impact and lowest upfront cost, such as setting up a COE and establishing a strong data management process. The next steps could then center on further sourcing more specialized talent and investing in analytics models and tools. Just two to three months of a learning-by-doing approach, starting with a small number of initial use cases requiring only existing capabilities, is often enough to build analytical muscle and achieve real impact—conditions ripe for further scaling.
Corporate functions that are further along the journey have a different challenge. They can make progress by continually engaging with the business to refine and renew the list of high-value business questions that remain unanswered—and then sequencing the development of new analytic models based on their complexity. This cycle allows analytic teams to move incrementally to more complex use cases as they gain experience, and as successive data, technology, governance, and talent investments begin to pay off.
Advanced analytics has the potential to improve the efficiency and accuracy of G&A functions when implemented successfully. Although implementing analytics may seem intimidating for companies just starting their advanced-analytics journey, application of these analytics can be a relatively quick and manageable process.
