Not all investments in digital appear to pay off—a selective approach is essential when digitizing the enterprise.
Digital technology is transforming the financial-services industry, and banks face the challenge of fully digitizing their businesses. To do so, they must decide where to invest in digital and how to justify these investments amid rising IT budget pressures.
Our research on the state of digitization among financial institutions in Eastern and Western Europe suggests where digital investments are best placed, most notably in automation of back-office processes and in sales-side analytics. In contrast, for example, investments in multichannel integration do not appear to have been as effective for these banks. We reached our conclusions by measuring factors across four dimensions: first, the level of digital enablement provided by IT with respect to end-to-end automation of processes, sophistication of digital channels, and analytics for decision making; second, the level of IT spending; third, the maturity of IT practices; and fourth, the bank’s level of overall cost and profitability.
By correlating the cost/income ratio with the level of digital enablement, we found that banks’ profitability is related only to specific areas of IT digitization. While correlation does not necessarily imply causation, it is interesting to see that more profitable banks have been investing in a few common areas. The areas with the highest correlation with profitability were product back-office automation, digitization of document management and automation of credit decisions, and big data analytics applied to sales campaigns. The profit margins of banks with high levels of digital enablement in these areas were, on average, twice as high as the profit margins of other European banks.
Consequently, while keeping in mind that each bank’s situation and investment cycle is different, we can formulate hypotheses about where IT digitization investments will provide the most bang for the buck (Exhibit 1).
In contrast, we identified several areas of digitization that do not show much correlation with profit margins. Among them were multichannel integration and sales-dialogue support. Apparently, investing in these areas, on average, does not yield a clear payback. This does not necessarily mean that companies should not invest in these areas. For instance, investing more in multichannel integration may be advisable for banks whose integration between channels is poor. But the complexity of their architectures may cause an escalation of project expenses and delays, therefore reducing overall return on investment.
Financing digital investments
Considering the budget pressures in IT, where should the funds for investing in digitization come from? One option is to create spending headroom within the IT budget by cutting costs. Our findings show there are opportunities to save both on day-to-day IT operations and on application development.
We found that certain categories of IT management capabilities show a strong correlation with lowering spending on day-to-day IT operations. The correlation was particularly strong for cost control, rigorous project prioritization, advanced sourcing practices, and relentless standardization of IT infrastructure and application architecture. Banks that manage these areas well spend, on average, 41 percent less on day-to-day IT operations than banks that have self-reported deficiencies in these fields (Exhibit 2).
Contrary to expectations, we found that several practices in application maintenance and architecture, usually considered drivers of savings, were not directly linked to lower IT spending. For example, frequent software releases, strict adherence to release schedules, and a modular application architecture were uncorrelated with spending. This doesn’t have to mean that these practices are unnecessary. They may help smooth operations without being a strong differentiator for managing spending; furthermore, many banks have already achieved a high level of proficiency in these practices.
For application development, the capabilities we found to be most related to lower spending were effective demand management, centralized application-architecture governance, and use of agile software development. On average, banks with self-assessed high capabilities in these areas devote only 3.5 percent of the bank’s total expenses to application development, while banks with lower capabilities allocate as much as 8.2 percent of expenses to the area (Exhibit 3). What’s more, this difference in spending is not attributable to the elimination of discretionary projects. We found that banks with high capabilities in areas related to application development on average manage to invest 62 percent of their application-development spending on customer-facing applications, compared with only 47 percent for banks with lower capabilities.
When comparing application-development methodologies, we found that agile practices help control costs as well as delivery time. Banks that apply agile methodologies to less than a quarter of their projects deliver 70 percent of projects on budget and 55 percent on time. In contrast, banks that use agile in more than half of their projects deliver 96 percent of projects on budget and 79 percent on time.
However, our research showed that a number of project-delivery capabilities do not contribute to reducing application-development spending despite being considered best practice. In particular, extensive automation of software testing, formalized processes for business involvement, and documentation of business requirements for software are not correlated with application-development spending. These counterintuitive results may indicate that too much automation and too many formalized processes do not yield additional benefits.
Banks urgently need to digitize their businesses, but they should invest selectively in areas where recent research indicates the best payoff. To meet IT budget pressures, CIOs have many opportunities to fund investments by cutting the costs of day-to-day operations and application development.