On Friday, I shared some scary statistics about ERP implementations and noted that ERP gives many companies fits. Yet here's the thing: Many companies also simply can't function without ERP systems. So there's a strong element of "damned if you do, damned if you don't" with ERP.
It's good to see a company emerging relatively unscathed from an ERP implementation and seeing benefits accrue rapidly, as appears to be the case with Peet's Coffee & Tea. Its ERP project is chronicled in two parts on SearchCIO-Midmarket.com.
The company's reasons for implementing ERP, detailed in Part 1, will sound familiar to many companies considering ERP. Peet's was experiencing strong business growth, adding new sales channels and sales partners, and had multiple business systems, many of them duplicative and outdated. Peet's had little visibility into its supply chain. Selecting and implementing a system wasn't easy, and the articles hint at some of the challenges.
In fact, CIO Tom Cullen used a slide emblazoned with the words "70 percent of all ERP implementations fail" (one of the statistics I cited from a Panorama Consulting study) in his presentation to executives because he wanted them to be aware of the level of difficulty involved. I think this was a wise move on Cullen's part. When I wrote about Waste Management's $100 million lawsuit against SAP over a failed ERP implementation, I shared the opinions of several observers who agreed at least some of the blame fell to Waste Management. The company had unrealistic expectations that the system would solve all of its business ills and tried to take on an ERP project in the midst of other major business challenges -- not a great idea considering ERP's well-deserved reputation for complexity. Culllen's quote:
I really wanted to put a little fear into everyone to say that they were signing up for a project that a company of our size had never been through and it would be business-changing.
Some of the other things Peet's did right:
Involve business users in the design phase, and document everything. Cullen had the company's business groups map out their current processes relating to five key areas: order management, fulfillment, inventory management, procurement and manufacturing. The goal was to choose the "core of core" areas that needed to be scaled or otherwise tweaked for efficiency. This phase accounted for a year of the 20-month project. Documentation is important, says Cullen, because sometimes business units will ask for a lot of changes once implementation is under way.
Overhaul infrastructure, with a focus on integration capabilities. Peet's adopted a standard integration interface using Microsoft BizTalk and a new virtualized production testing environment for integration that would take place between back-end systems and the new ERP system.
Choose a system with an interface familiar to users. That was Microsoft Dynamics AX for Peet's because the company is a Microsoft shop.
Get an experienced partner to help. Peet's worked with a systems integrator that had experience designing ERP systems for the food and beverage industry and for companies with multiple sales channels. The SI also had worked frequently with Microsoft Dynamics AX, its chosen system.
Avoid customization as much as possible. SI Junction Solutions created an approach in which data from all of the company's sales channels are funneled into a consistent order-and-fulfillment process. Minimizing customization is important, says Junction Solutions' director of consulting services, because "more customization means a longer timeline, more testing, more costs and more likelihood that the project will fail."
Involve business users in training and education. Business unit leaders wrote their own training manuals and were involved in employee training efforts.
Another key to avoiding excess ERP customization, related in Part 2, was simultaneously retiring legacy financials, order management, inventory management, procurement and fulfillment systems and making them ERP modules within Microsoft Dynamics AX. Peet's decided to hold off a little longer on manufacturing because it relies on all of the other modules to work properly.