MoBT_38_Transitioning_to_standard_software_1536x1536_Original

Transitioning to standard software: Lessons from ERP pioneers

By Sanjay Kaniyar, Peter Peters, and Ulrike Vogelgesang

Banks and insurance companies are at an inflection point with their aging proprietary IT systems. A move to standard software may allow them to save time and costs and create new digital services.

Since the late 1990s, many production and manufacturing companies have replaced their proprietary data-management and control systems with standardized, packaged software, such as enterprise-resource-planning (ERP) systems and customer-relationship-management (CRM) databases. The switch has allowed them to automate important business functions—for instance, implementing one-touch customer billing or automated supply-chain planning—and reap significant cost advantages from shared services.

In banking and insurance, however, the use of proprietary software systems is still the norm. For companies in these sectors, upgrades to core systems are particularly lengthy and risky—because they typically involve many complex, heterogeneous products and decades-old IT systems, and they require input not just from multiple internal stakeholders but also from external parties, including regulators, before anything can change. Moreover, the market for standard software products that are specific to these industries is still emerging.

A confluence of technology and business trends, however, may finally prompt insurance and financial-services companies to make the leap, not the least of which is increasing customer demand for online products and services. Companies in all industries are experimenting with two-speed technology platforms—rapidly developing innovative website and mobile applications on the front end that facilitate better interactions with customers while continuing to run standardized legacy systems on the back end to ensure data security and reliability.1

Banks and insurance firms would like to do the same, but their proprietary platforms are proving too inflexible for creating new digital channels or user interfaces. Because such systems are generally built piecemeal and patched together over time, they typically cannot enable the end-to-end data and process f lows and automation that companies need to provide fast, reliable online services to customers.2

Through our work with companies in a range of industries, we have identified five success factors for switching from proprietary to standard software systems. Specifically, executives leading the change effort will need to pay attention to the technologies required, the impact of transformation, the composition of teams, the timing of system implementation and rollout, and the organizational transparency required to successfully shed homegrown systems in favor of standard systems.

In this article, we consider the software standardization challenge through the lens of the insurance industry, and we elaborate on the five success factors. By paying attention to them, insurers and others can shorten the adoption curve, accelerate proposed systems changes, and capture the same kind of cost and service advantages that production and manufacturing companies are beginning to enjoy.

Standard software and the insurance industry

Change is a constant in insurance, as it is in every industry. Multiple brands and business segments are consolidating, new regulations have been introduced, and multichannel business models are in demand. Today’s customers have come to expect digital services from their insurers, and many companies are trying to respond with web-based sales channels and innovative insurance products—for instance, automotive-insurance-policy premiums that are based on the use of in-vehicle telecommunications devices (telematics).

To support digital products and services, insurers require business and IT architectures that emphasize stable processes, accurate and reliable data, and a short time to market. But most insurers are still saddled with convoluted, outdated legacy systems and a shortage of programmers who have the necessary design and coding know-how to manage the older technologies and synchronize them with newer ones.

One midsize US life-insurance carrier sought to introduce new life and retirement products online, but its customized—and now outdated—policy-administration system was unfit to support this effort. Another major life insurer in Europe was spending about 10 percent of its total IT budget to maintain proprietary back-end systems so that they could keep pace with changing legal and other mandatory requirements. The company had nothing left to spend on new or enhanced features, and only a handful of programmers on staff were familiar with the proprietary software. As a result, the insurer had trouble meeting its regulatory requirements and had insufficient IT support for many back-office processes. Service to customers was limited, and operating costs were high.

Industry experts have long predicted that insurance companies would move toward standard, packaged software to address those and other issues. But the sector has been slow on the uptake for several reasons.

High levels of complexity

Replacing core IT systems is especially onerous for insurance companies because they have aging IT infrastructures (many are more than 20 years old) and a highly heterogeneous, highly complicated product mix. Product lines might include automotive, pension, or health policies, and processes may swing from underwriting home loans for commercial risks to handling healthcare claims in cooperation with third parties. Underlying such insurance products and processes more often than not is a heavily siloed and fragmented application landscape. For example, life-insurance policies need to be serviced for more than 50 years, and so the information-storage and support requirements for these products are more complicated than they are for automotive, property-and-casualty, or health-insurance products, which typically have shorter spans of service activity.

Immature software market

Manufacturing and production industries have access to software packages that take their business contexts into account. There is software for procurement and materials management, for instance. Vendors have even established industry-specific “process languages” for companies in chemicals, pharmaceuticals, and utilities. By contrast, insurers are still seeking—and by and large have been unable to find—standard software packages that address the challenges they face in policy and claims management and that contain features critical to their ability to meet various business and regulatory requirements. End-to-end IT products are also scarce. In many smaller countries, for example, insurers must decide whether to engage global vendors that can provide modern platforms but lack country-specific product or process features, or local providers that are well adapted to local needs but use dated technologies and provide only a limited set of features. Questions about technology “lock in” and the long-term viability of vendors remain concerns as well, given the life span of insurance products and the continuing trend toward consolidation among software providers.

Budget constraints

Like companies in other sectors, insurers face strong margin pressures and are constantly reviewing their IT spending. Meanwhile, the current trend toward digitization requires fast, targeted products and platforms. There is little appetite, therefore, for core-systems replacements that can take up a large share of the IT group’s resources and management’s attention over a multiyear period. After all, the full financial benefits of IT-system transformations are often seen only after the end of a CEO’s tenure. (See sidebar, “Insurance 360º survey: A case for standardization.”)

Inertia

For a long while in the insurance industry, income from policy premiums was high and relatively predictable; this allowed insurers to continue relying on their legacy systems, even if they were costly to maintain. But the advent of the financial crisis and the increase in price transparency (through insurance aggregators and websites, such as Confused.com and CHECK24) have changed all that.

Standard software would appear to be an imperative for insurers, but the time and money required to make the change are limited. The software-vendor market is emerging, but the track record is scanty. Standardization is not a panacea, of course, and every insurer has to find its own way through this rocky terrain, but core-system-replacement lessons from the manufacturing and production industries could make the journey shorter and significantly reduce the risks.

Lessons in standardization

Manufacturing and production companies have spent most of the past two decades standardizing their IT backbones—and many of them are still refining their systems to accommodate digital trends. But 20 years is a long time. How can insurers and other financial institutions avoid or at least curtail this lengthy adoption curve? We have identified five critical success factors:

Technologies

It’s generally true that the most commonly used software sets the standard. In recent years, companies such as SAP and Salesforce.com have taken leadership positions in the markets for databases, classical ERP systems, and software-as-a-service options.3 Meanwhile, software companies such as Guidewire Software and Accenture Duck Creek were among the vendors with the highest numbers of new installations of property-and-casualty policy-administration systems in the industry in 2013, according to Gartner.4 However, before selecting from among market-leading and other technologies, insurance companies should set some selection criteria that may include provider size and number of customers served, the provider’s functional and technical competence, or inclusion of industry-specific features in provider’s products. Insurers can use information on vendor websites and in trade and mass-media publications to conduct this research.

Transformation

Any transition to standard software should not simply be considered a software-implementation project. It also needs strong leadership from the business side and the IT group, a clear accounting of potential business value, and a strong focus on the principles of change management. Indeed, it’s critical to get corporate leaders to see the explicit connection between systems and business requirements; the likelihood of project success increases, not just from coming in at budget but also from speed to market and overall impact created for customers and the company.5 Specific to insurance companies, transformation projects should start with a clear understanding by both business and IT leaders of the tangible improvements that brokers and agents will receive from the systems update (comprehensive support in their day-to-day underwriting and claims-management tasks), that customers will receive (they can access insurance products quicker and through multiple channels), and that the organization will realize (IT and operating costs can be reduced).

Team

Success in large-scale IT projects requires all hands on deck, but it can be easy for companies to overestimate the availability of the talent required in their organizations. Many insurance-company IT transformations have been thwarted by the lack of well-versed actuaries and data analysts. Companies also must invest in good program managers and information architects, either by hiring them directly or by ensuring that a systems integrator provides them. Qualified people from the business side should be staffed on the transformation project full-time; some companies have placed rising young stars on these projects, where the lack of institutional baggage and an entrepreneurial perspective can be particularly important.

Timing

Avoid projects longer than your business cycle (a maximum of two to three years). The legend that companies face a “once-in-a-lifetime opportunity for change” is just that—a legend. In companies’ shifts away from proprietary software systems, frequent projects of shorter length are often more successful, because they are less complex and companies can focus their full energy on each project for an acceptable amount of time. Insurance companies may want to concentrate on only those elements of the software transition that will most affect critical processes; other functions can be targeted at a later date. Many production and manufacturing companies that have been through this transformation would likely argue that, in retrospect, their rollout schedules for new standard software were too slow. Because they had the time, they took the time—continuously tweaking the scope, functionality, and technologies associated with the rollout and then getting mired in those changes. Many projects came to a complete standstill and had to be relaunched as a result.

Transparency

Implementation can succeed only with effective governance. Surprisingly, companies can forget some of the basic requirements when it comes to standardization projects: creating transparency, emphasizing the discovery and mitigation of issues and risks, paying close attention to project scope, making decisions quickly with complete buy-in and accountability from a strong steering committee, and rigidly managing the consequences of those decisions. These factors are often neglected or compromised as fragmented organizational interests assert themselves and power plays occur. It is therefore all the more important to have a steering committee that accepts accountability for project success, understands its role as the decision maker (not a reporting body), and is empowered to drive the project forward despite organizational resistance.

In our experience, insurance and other companies that have applied these five principles have often completed their integrations of standard software within two to three years. They have then moved on to capture the opportunities created by the introduction of common processes, data, and systems in the digital world. The results have been impressive. For example, a logistics company is using “real time” reporting to make more accurate planning decisions. One insurer was able to reduce its overpayment of claims because of the enhanced pricing and underwriting capabilities it was able to realize by using standard software—the system now incorporates new external variables for the underwriting pricing model. Another insurer has managed to substantially reduce costs associated with its new policy-administration system by following the standards recommended by the software maker rather than customizing the software extensively. The company estimates that it is now 30 percent more efficient in rolling out new products and services because of the rules-based foundation its standard software provides.


Ten years from now, IT architectures in the insurance and banking industries will be substantially different than they are today—in both composition and desired functionality. The financial institutions and insurance firms that start now to reevaluate their proprietary systems and reinvest in upgrading their IT architectures will be well positioned to capitalize on multichannel marketing, digital-product innovations, straight-through claims processing, and other lucrative digital-enabled opportunities.

About the author(s)

Sanjay Kaniyar is a principal in McKinsey’s Boston office, Peter Peters is a principal in the Düsseldorf office, and Ulrike Vogelgesang is a senior expert in the Hamburg office.

Article - McKinsey Quarterly

Culture for a digital age

Article - McKinsey Quarterly

The CEO’s guide to competing through HR