Seven Steps to a Successful ERP or CRM Launch
Keep your project team on track.
Yesterday, I wrote about the growing interest in two-tier ERP, in which companies maintain a central ERP system but supplement it with an ancillary system or systems in some of their business units, often subsidiaries in another country and/or units that have regulatory or process requirements that differ from HQ.
While this strategy can help contain ERP costs and increase process efficiencies, not surprisingly, it can be tricky to implement. Writing for IndustryWeek, Dinesh Mohan, an industry principal for Enterprise Solutions with Infosys Technologies, offers some tips for selecting and implementing a two-tier ERP strategy. The primary factors to consider, as outlined by Mohan:
- Corporate structure (e.g. centralized versus decentralized governance)
- Business needs and priorities, including timeline constraints that drive the ERP implementation
- Existing IT landscape, including legacy line-of-business applications
- Organizational change management challenges
- Corporate IT strategies and long-term roadmap for optimizing software licenses and hardware costs
- Maturity of IT organization and infrastructure in both HQ and distribution units
- Availability and extent of budgets for the ERP solution
After considering these factors, companies can choose between three options: a single ERP with a single instance, a single ERP system with multiple instances or multiple ERP systems with multiple instances.
Mohan concludes the single system/single instance model tends to work best for medium-sized enterprises with a centralized management and governance structure. The single system/multiple instances model works best for larger enterprises with not many operations outside HQ. Larger enterprises with a more varied portfolio of subsidiaries tend to be the most appropriate candidates for the multiple ERP/multiple instances model (referred to as a two-tier or hub-and-spoke approach).
There are risks associated with the two-tier model, some of which Mohan lists:
- Lack of standardized business processes, which can reduce efficiencies and increase costs
- Added software licensing, maintenance and hardware costs
- Need for more diverse skills among IT personnel, which can increase dependence on system integrator partners, also driving up costs
- Increased difficulty in maintaining data quality, which may entail the deployment of data warehouses and/or data marts
In a recent interview, Paul Turner, senior director of product marketing for cloud-based ERP provider NetSuite, told me some of these issues, specifically the second and third ones on Mohan's list, could be mitigated by opting for a software-as-a-service (SaaS) ERP system. SaaS ERP seemingly would fulfill several of the characteristics Mohan lists as desirable for "spoke" systems:
- Provides reasonably deep integration across general ledger, accounts payable, accounts receivable, fixed assets, sales, purchases, inventory and other centralized functions
- Can be easily deployed in most industry verticals
- Utilizes a simple interface to reduce need for user training
- Minimal requirement for new IT infrastructure and added staff resources
- Integrates with legacy applications and with productivity software, data warehouses and other necessary connection points
- Has enough scalability and flexibility to support the larger enterprise for the foreseeable future
Mohan also provides three suggestions for implementation of two-tier systems. As is so often the case, these suggestions would be good for any deployments but are especially important to reduce some of the particular challenges of the hub-and-spoke approach:
- Carefully select the system integrator partner. Mohan says those with experience in installing large ERP systems are the best candidates, thanks to their experience with change management, process standardization and other aspects of ERP rollouts.
- Pay special attention to project management, selecting the right project committee members and ensuring appropriate control, structure and communications mechanisms are in place.
- Establish data management rules upfront, including acceptable sources of master data, consumers of data and anyone who possesses rights to modify data.