What makes for CRM system success — Or failure?

Abstract

Although customer relationship management (CRM) has been one of the fastest growing businesses of the new millennium, critics point to the high failure rate of the CRM projects as evidenced by commercial market studies. The purpose of the study is to investigate success and failures of CRM system implementations. We found that the scope, size, complexity and duration of the CRM projects seem to vary quite significantly across firms. Poor planning, lack of clear objectives and not recognising the need for business change are the key reasons for CRM failures.

INTRODUCTION

Since the late 1990s Customer Relationship Management (CRM) has been one of the fastest growing businesses and energetically debated topics among practitioners and academicians. Companies have invested or are planning to invest huge amounts to implement CRM strategies, tools and infrastructure in order to win the battle in the increasingly competitive economy.¹ As a result, the growth in demand for CRM solutions has been increasing. Gartner estimated that the market for CRM software exceeded $7.4 billion (£3.6 billion) in 2007, up 14 per cent from 2006.² Forrester estimated moderate growth in the CRM industry through 2010. Forrester suggested that worldwide revenues for CRM solution providers reached $8.4 billion (£4.08 billion) in 2006 and would continue to grow to $10.9 billion by 2010. Forrester anticipated that overall CRM spending will remain steady, with services taking an increasing share of vendors' revenue.³

Despite the enormous growth in the acquisition of CRM systems in the last ten years and widely accepted conceptual underpinnings of a CRM strategy, critics point to the high failure rate of CRM implementations as evidenced by commercial market research studies.⁴ In an international survey of 1,337 companies who have implemented CRM systems to support their sales force, CSO Insights has estimated that only 25 per cent reported significant improvements in performance.⁵ According to a Gartner survey, about 70 per cent of CRM projects resulted in either losses or not bottom-line improvement in organisation performance.⁶

Manufacturing, logistics, store/branch operations and the public sector have all produced great examples — of both success and failure. For years, change programmes involving a big change to systems, or even totally new systems, involved additional risk. It is no surprise — the more the change the greater the risk. CRM systems success and failure have attracted a lot of publicity in recent years, but in the last quarter century, there have been many other management areas where the issue of systems success and failure has attracted the same interest. The reasons for success and failure turn out to be pretty closely related. Successful systems-supported change projects observe change management disciplines, unsuccessful ones do not. Successful projects are planned carefully, with all aspects covered, all the right people involved, and the company and any external suppliers (consultancies, systems companies, business partners) work well as a team. The opposite applies to failed projects. In the middle lie the many partly successful projects that achieve some but not all of their objectives, probably at a higher cost than initially planned.

One of the central themes of most studies of success or failure in systems development and implementation is whether the project slipped badly, so that it was completed well after the original deadlines. In our experience, however, CRM systems seemed to be different, in that they usually support a large change in how a company works with its customers, a change that involves learning by the company's staff, its system suppliers, the consultants (if any — and internal or external) helping plan and manage the projects, the company's business partners (eg marketing communications agencies, call centre capacity suppliers, fulfilment houses) and not least, its customers — as the company tries to create a new way of working with them, and sometimes even a relationship. These CRM systems often had several stages of development, with one or two years of pure development, followed by several years of implementation and iterative improvement. Our interests, therefore, lie in exploring factors that contribute to successful CRM implementation as experienced by users in the private sector.

WHAT IS A CRM SYSTEM?

Payne and Frow point out that there is a lack of consensus on the definition of CRM.⁷ A narrow-based definition of CRM often contributes to the failure of CRM projects when an organisation views CRM from a limited technology perspective or undertakes CRM in a fragmented way. A CRM system is a technology-based business management tool for developing and leveraging customer knowledge to nurture, maintain, and strengthen profitable relationships with customers. Thus, a CRM system is an essential part of a global CRM strategy which emphasise creation of shareholder value through the development of appropriate relationship with key customers and customer segments.⁸ The underlying premise of CRM is that firms create customer knowledge in order to (1) effectively segment customers, (2) develop and maintain long-term relationships with profitable customers, (3) determine how to handle unprofitable customers, and (4) customise market offerings and promotional efforts.⁹ Through the creation and better utilisation of customer knowledge, the customer relationship and customer loyalty should improve.

As suggested by Raman et al. CRM system applications can be categorised into operational or analytical.¹⁰ Operational CRM aims to reduce operating costs while enabling these functional areas to provide a higher level of value to customers. It contains all applications directly in contact with the customer (eg front offices). Operational CRM involves sales force automation (SFA), marketing, and customer support with a view to making these functions more efficient and effective. For instance, a CRM system may guide a salesperson through identification of customer-related information when making a sales call. This information can immediately be connected to back offices or transferred to other functional departments (eg manufacturing, finance, and logistics) and communication channels in order to identify and provide the customer with a market offering that provides value.

Analytical CRM concerns the technologies that aggregate customer information and provide analysis of the customer data to improve managerial decision making and actions. It is based on technologies such as data warehousing and data mining. Ideally, the customer database should be accessible from all relevant departments such as sales, customer service, and marketing. Analytical CRM forms the basis for planning and evaluation of marketing campaigns and assists cross and up-selling functions. To implement CRM successfully, firms must combine physical resources (eg computers and technological infrastructure), informational resources (eg customer databases, salespeople's call records, customer service interactions) and organisational resources (eg customer-oriented culture, information-sharing routines) to enhance relational resources (ie relationships with customers) so as to improve a firm's competitive position. CRM implementation success may be defined as occurring when a CRM system helps a company profitably deliver market offerings to customers that (1) provide value to customers — possibly at a lower cost (relative to competition), (2) provide more value at the same relative cost (relative to competition), or (3) provide more value at a lower cost (relative to competition).¹¹

SUCCESS AND FAILURE OF CRM SYSTEM IMPLEMENTATIONS

Payne and Frow argue that successful implementation of CRM programme depends upon four critical factors: (1) CRM readiness assessment, (2) CRM change management, (3) CRM project management, and (4) employee engagement.¹² A CRM readiness assessment is an overview audit which helps managers to assess the overall position in terms of readiness to progress with CRM implementations and to identify how well developed their organisation is relative to other companies. CRM change management involves strategic organisational change and cultural change. Senior level understanding, sponsorship, leadership and cross-functional integration are clearly critical in a complex CRM implementation. CRM project management requires forming cross-functional teams of specialists who manage the enterprise's CRM implementation programme. Successful CRM projects deliver against the CRM objectives derived from the corporate objectives and support the overall business strategy. Finally, employee engagement comprises support and commitment of the employees to CRM projects. Increasingly, firms recognise the significant value their employees contribute to the business, which extents well beyond the basic fulfilment of core duties. Companies cannot develop and operate appropriately customer-focussed CRM systems and processes without motivated and trained employees.

Various arguments have been put forward for the failure of CRM systems. Day suggests that main reason for CRM project failure is the lack of strategic planning prior to the implementation of CRM.¹³ Maselli found that the reasons for failure of many CRM initiatives ranged from technological implementation problems to a lack of organisational integration and customer orientation.¹⁴ Jain et al.¹⁵ state that most such failures are attributed to poor design, planning and measurement of CRM projects. They comment that capturing the wrong customer information, unclear goals, inappropriate selection and use of technology, inability to integrate people and processes and use of misleading metrics or improper measurement approaches are the major barriers in implementing and managing CRM projects. According to Kale,¹⁶ the seven deadly sins for unsatisfactory CRM outcome are: (1) viewing the CRM initiative as a technology initiative; (2) lack of customer-centric vision; (3) insufficient appreciation of customer lifetime value; (4) inadequate support from top management; (5) underestimating the importance of change management; (6) failing to re-engineer business processes; and (7) underestimating the difficulties involved in data mining and data integration. He states that most executives are not even aware of these issues, even though they could spell disaster for their careers and for the company. Others argue CRM failures are heavily influenced by the firm's lack of ability to integrate CRM technologies into its functional processes.^{17, 18}

RESEARCH METHOD AND SAMPLE

We approached about 90 of our contacts we considered had direct and personal experience of CRM developments and deployments in the last ten years. They included supplier-side contacts, with knowledge of many client projects, but also client-side contacts personally involved in one or more CRM projects during that period. Some of the latter had moved jobs and so experienced more than one CRM implementation — particularly as they are likely to have been hired across from one company to another because of their valuable experience. We received nearly 40 responses, showing the accuracy of our targeting and the responsiveness of our contacts. We asked them a few open-ended questions. The following section presents our findings.

FINDINGS

Different industries and users have different priorities and varying amounts of funding to invest in CRM. Companies with very tight profit margins or limited funds tend to partner where they can (eg shared loyalty schemes, outsourced database management and call centres) and to use small ‘pay as you go’ development stages, while keeping a consistent vision and objective in mind. Many of the largest scale projects were in the financial services industry in the early years of this century. Banks, insurers and others usually used packages such as Siebel and Chordiant at the heart of their project, but some developed their own systems. These companies were moving — or had already moved — from managing customers mainly in the branch or direct mail, to managing them in the contact centre and over the web. Here, it is arguable that the main reason for the investment was to cut the cost of managing customers, and to enable the company to move into new product markets very quickly (eg from banking to insurance or vice versa). At the same time, telecommunications companies were investing in large CRM projects. Some did it as a ‘Big Bang’ replacement for an existing system which was focused entirely on subscriber management, including billing. Their new systems enabled them to manage churn and up/cross-sell much better. In telecommunications, particularly in mobile telephony and now broadband, reducing churn is a critical target, while increasing customer value by up-sell and cross-sell is the ‘cream on the cake’.

Many of these companies have sustained their CRM systems focus over time, although these investments may or may not be considered part of a long-term CRM strategy. For example, one major international bank defines its CRM systems as the marketing databases and campaign management and considers distribution channels to be a separated (but linked) systems investment area.

Industries with tighter margins (eg forecourt sales) appear to have invested more cautiously but have usually sustained this over a long period to achieve substantial effect. With limited funds we saw far greater focus on challenging proofs and pilots, also a strong management process for reinvesting early savings to create a ‘pay as you go’ approach. This approach is increasingly used within the public sector, where ‘lean’ initiatives can generate savings that can be immediately reinvested for additional productivity gains.

The most recent CRM programmes, for example some in local government, have clearly benefited from previous CRM systems experience of suppliers, including considerably reduced implementation times and lower risk levels for comparable size programmes.

All these companies and organisations had very different levels of success. Their success was determined mainly by the relationship between the complexity of the system and the speed and phasing of its development and roll out. A CRM system is not just customer interface software.

CRM PROJECT TYPES

Our research was broad and allowed respondents to define and describe ‘CRM systems’, as we found that these varied widely by market, industry and company. We received a very broad list of CRM project types, demonstrating that there is perhaps no common definition of CRM systems content. This list summarises the systems technologies deployed by companies in earlier years. Projects may include just one of these technology items listed, or a small number, or very many of them when included in long-term and complex programmes.

Customer data and analytics

• — real-time client data files, marketing databases, data warehouse and data marts;

• — customer segmentation, propensity modelling, profitability analysis and other analytics.

Marketing and campaign management

• — marketing campaign management, including via email;

• — marketing resource management (MRM);

• — event and survey management.

Distribution channels

• — contact centres for service, or sales;

• — SFA and key account management;

• — branch or store merchandising and support;

• — self-service via web, mobile or voice recognition;

• — field service support;

• — partner management.

Many other topics might now be considered part of evolving CRM programmes. These innovations update the CRM concept through improved research, analysis and contact management in support of profitable growth and an improved customer experience

Research and analytics

• — Real-time decision making and ‘next best offer’ recommendations;

• — customer experience management and customer journey mapping;

• — mobile polling for real-time decision and near decision insights.

Distribution management

• — online chat and voice over internet (VOIP), click to call back and email call back;

• — case management and workflow management;

• — collections and debt management;

• — outsourcing and work load balancing.

Integrated regulatory compliance

• — treating customers fairly (TCF);

• — anti-money laundering (AML);

• — know your customer (KYC);

• — complaints escalation and management (including ombudsman processes);

• — risk management and pricing;

• — Other industry-specific regulatory and compliance issues (Telco, Pharma, etc).

Profit and portfolio management

• — profit assessment and management;

• — fraud detection.

The activities commonly found in CRM projects include:

• — contact centre for sales and/or for customer administration and service (and in some cases this will include with some self-service via IVR or voice recognition, although some projects were focused on adding this to an existing contact centre);

• — adding additional capacity, for example, a new contact centre, sometimes but not always based on an existing system;

• — web-based customer management (which implies a move to more of a self-service model);

• — integration of customer management across several channels, for example, branch, contact centre, web, back-office;

• — customer database development;

• — data warehouse or data mart — mainly for analysis and planning;

• — marketing campaign and resource management;

• — integration of any existing CRM system with core/legacy systems;

• — support of mobile or field, sales or service, staff.

So, when someone talks about ‘their CRM project’, they could mean any or all of the above. Each function usually demands a big additional development, customisation and/or integration effort, depending upon the choice of software.

THE EFFECT OF SIZE AND COMPLEXITY

The number of CRM system users is not by itself a major factor in the cost of development, unless there are different types of user with different business requirements, for example, field sales versus contact centre sales. The number of users usually affects cost mainly through the number of software licences needed and recruitment and training costs.

‘BIG BANG’ OR STAGED DEVELOPMENT?

Complexity and scale, however, both affect the time it takes to develop and roll out a CRM system of given functionality. ‘Big Bang’ approaches, the creating of large-scale projects with high levels of functionality, are not uncommon. However, although this approach might seem risky, it can be greatly de-risked by not tightly coupling too many elements of the programme. For example, a company which wants both to improve dramatically the quality and scope of its customer database and the quality of interaction with customers in the contact centre does not have to premise its investment on using the new data through a new customer interaction system. Some or all of the new data can be planned to be used with existing customer interaction system, even if full exploitation of the data has to wait until the customer interaction system is completed. So, although we know that in information technology projects, project risk increases with project size, the risk can be reduced by having several interlocking projects where attention has been paid to developing a ‘Plan B’ for deployment of whatever has been successfully developed even if there are slips in other parts of the programme.

Many companies reduce risk by staging the development of any particular functionality or group of functionalities, so that benefits begin to be delivered early on, and continue to be developed from deployment experience. This also reduces the political risk of such programmes. For this reason, many suppliers of CRM systems and consultancy have created CRM programme planning and deployment methodologies that support accelerated early development stages. At the heart of such methodologies must be a strong CRM programme planning methodology that is understood by senior management as being the central focus of their governance for their CRM improvement efforts. The figure below, from Customer Essential, illustrates such an approach.illustration

The risk of a Big Bang approach is that all the requirements for development changes are gathered at the beginning of the project and applied in one continuous development period, before ‘waterfall’ testing takes place and subsequently deployment. The waterfall model is a sequential process for developing software in which development flows steadily step by step, like a series of small waterfalls, through requirements analysis, design, development and integration, testing, implementation and maintenance. Although there are many arguments about the pros and cons of following this model, it seems that large-scale projects that use this approach tend to have a high risk of failure, longer time to benefit and are also less adaptable to continuous learning during the development and deployment process. As a result the development effort invested in some functions may be wasted. The waterfall model is associated with another systems development idea — that of Big Design Up Front, in which system design is agreed early on in the project so that the systems development effort can focus consistently on the required design. Of course, there is no perfect model for systems development. Indeed, the waterfall method is often criticised because in the real world requirements do change, as systems frequently need design changes before their development is finished. As a result it is usually not possible to have smooth, step-by-step progress from design to implementation. Instead, an iterative approach often proves much more effective at delivering appropriate business value in stages and reducing programme risk. We believe that companies which think hard about the options for their CRM programme phasing and systems development methodology and consider them as related aspects of the same task are more likely to get the CRM systems they want.

‘OUT OF THE BOX’ OR CUSTOMISED?

The appropriateness of the software chosen to implement a particular functionality is also important. Some packages are suitable for implementation ‘out of the box’, provided that the company is prepared to change how it works to fit with what the software designers conceived. Others require significant customisation or development. Of equal importance is the choice of systems integrator (unless the work is done in-house). Systems integrators with good track records of success (and this covers not just development, including experience with integrating the particular software products chosen, but also implementation and subsequent support) are more likely to deliver a successful outcome than those with questionable records in this area.

HOW LONG?

We asked our respondents how long it took to complete their CRM programme. Given what we have said above about the content of a programme, it is no surprise to see that the answers varied dramatically according to this. Some respondents gave examples of initial development periods as short as 16 weeks from ‘go’ decision. Many others were developed within a year from decision date. This speed was usually achieved because the CRM programme was defined as development of a single function (eg marketing campaign management), with some integration to existing systems and data sources (even though this might effectively be the first stage of a multi-stage CRM programme). In other cases, it was achieved because the programme consisted of minimal customisation of an existing CRM package with very limited integration with existing systems or data sources.

Projects that took between two and four years involved multi-function developments. These are the majority of what we would call ‘comprehensive’ CRM programmes. However, when we analysed the responses we received, we saw signs that these programmes are now being completed a year or so more quickly than they used to be. We think this is because of the increased maturity of CRM software and the wider availability of CRM services, skills and proven development methodologies.

Projects that took five or six years appear to be, on the one hand, very large projects or where investments are deliberately stretched out over a long period, or on the other what are known as ‘troubled’ projects, which run well over time and budget compared to initial investment estimates.

When analysing recent primary and secondary research we found that CRM projects and programmes grouped by duration time as follows:

Up to one year: These were mostly single-phase or single-focus projects, usually including limited or no customisation, integration work or business change.

Two, three or even four years: These multi-phase or topic programmes were of various complexity, they often included substantial development, customisation, integration and/or associated business change.

Five, six or seven years: These were often very large-scale programmes (eg IBM and Citigroup global rollouts), while many were ongoing incremental programmes (eg Lloyds TSB customer insight project or BP forecourts project). Some projects in this category would have been classed as ‘troubled projects’ at some time, where a major redefinition and reset would have been required at some stage.

In addition to the above timescales, we also identified programmes that had no specific scope or end date. This is not because they were poorly planned, but rather because the company is on a long ‘CRM journey’ and committed to making a series of improvements to its CRM capability as changes take place in its distribution channel strategy, customer base, competition and the like, as customer management technology changes, and as the company learns what works and as their customer learn to be responsive (or not!) to the new customer management approach.

WHAT HAPPENS WHEN THINGS GO WRONG

A significant finding from this research and our own experience is that many issues remain unrecognised for far too long after they are first identified. While this can be the result of misunderstanding or a deliberate cover-up, it seems in most cases insufficient assurance enables the severity and impact of issues to remain undiscovered by the programme for far too long. By the time issues become recognised it is often too late to address them with tactical and immediate corrections, as a result the programme is often proven to be over time and budget and the downstream programme can only be rescheduled. Unless the delivery contract with external suppliers protects the budget, rectifying any issues will take the programme costs beyond the spending plan, forcing later deliverables to be cut unless more money is available. Programme managers often return to management for approval of more time and budget, as overruns become recognised, but at each request the confidence of management in achieving an on time, on budget and full benefit deployment is reduced. If delays and overruns continue, management confidence is lost and the programme is cut severely to focus on minimal deliverables. Such programmes may be gradually rebuilt over time, with additional budget, as confidence in the programme returns.

During the period of delay and recovery user departments often suffer budget freezes until the programme starts to deliver again, as the programme was probably put in place to provide for the primary needs of the user department. Even where requested budgets are related to new investment areas not covered by the programme they may still be held back until the programme makes a little progress. During this time the business can be seriously affected by the delay. In 2007 a loans company collapsed from the costs of continual parallel running of an old and new system for a few years after the new system deployment went continually over time and budget.

The additional costs and lost benefits are not only those of the programme but can be much wider. Business benefits are often delayed or lost (eg customers not retained, cross-sell not achieved, new business areas or propositions not launched, new markets not entered). In one case a company almost completely withdrew from a profitable B2B marketplace where the executives determined that there was insufficient capacity to operate there while recovering from a CRM ‘troubled project’ in their priority B2C business. In some situations projects are completely scrapped and financially written off, usually following a tough assessment of progress by the board. Future programmes in such businesses usually receive far more risk management attention at all stages from justification to deployment.

The keys to good governance are (mostly) well known, but often not applied. The role of the board is central and often underestimated, at all stages from justification to deployment and through the achievement and measurement of resulting benefits. Key factors include:

• — board involvement at all stages;

• — clear objectives and benefit outcome rather than transactional measures;

• — staged developments, deployments and benefits realisation efforts;

• — programme board with regular meetings and ‘clout’, key suppliers can be included;

• — working board and self-assuring programme streams, with project office support;

• — a business and systems architectural perspective;

• — external programme assurances at key checkpoints, used constructively rather than as a diversion of resources.

Many problem areas are associated with supplier selection, contracts and management. They include:

• — Executives who personally override ‘due diligence’ results and considered recommendations…

• — Supplier and solution decisions made by internal departments or contractors, rather than those that will ‘carry the can’.

• — Lack of business involvement in decisions.

These problems can be overcome by

• — tough profiling and appraisal of suppliers, including taking up of references for challenged projects, not just the references given by the supplier;

• — alignment of contract objectives, both commercial and personal;

• — fixed price, staged payments, time and materials or shared reward;

• — improved communication with suppliers.

THE KEY QUESTION

In the end, the most important question is — did the system work? The Gartner Group researched whether CRM programmes met expectations and identified that just over half were considered unsuccessful, though many delivered substantial benefits. In Gartner's view, too many companies failed to set clear objectives or to recognise the business change needed and often considered the CRM programme to be just a systems programme. We agree with this conclusion, as did many of our respondents.