The week after go-live resembles a newborn colt just getting up on its legs. Logically we couldn't measure our success by one statement or even one wobbly week. But after 26 months and tens of thousands of hours of analysis, development, implementation and training, our customer's comment made me reflect on what we have accomplished.
Two years ago, we decided to replace our 20-year-old legacy system. It was increasingly unable to meet the demands of our customers and vendors, and it was in danger of being orphaned by new owners of the company.
Its replacement had to allow us to cut costs, react profitably to business opportunities and improve customer service. We needed a vendor partner that would endure.
Getting Some Things Right the First Time
We did a lot of things right.
Starting with 22 candidates, we whittled the options down to two finalists in just a couple of months. It took us another six months to choose our vendor. The most critical factors were the candidates' corporate culture, their industry knowledge and the functions integrated into their product line. Functionally, the final two companies were very competitive with each other. Cost, always a factor, would be a matter for negotiation.
Balancing our successes and setbacks, we still believe we made the right decision in choosing Intuit Inc.'s Eclipse. The vendor's industry knowledge was as strong as we'd hoped. Our ability to conduct business electronically with our trading partners has already provided significant labor savings. We can now move from order creation to invoice payment with minimal human intervention. Complex transactions, many unique to the hard-goods industry, work as advertised. And on several occasions when we needed to meet the new demands of our vendors, these vendors said, "We've worked with [Intuit] Eclipse on this. No problem."
Even this early in the game, we have seen the benefits of fully integrated products. Document imaging, workflow queues and radio frequency-wireless warehouse management functions are all completely integrated into the software. Having a single database and one code base prevents errors. The similarity of screens and functions improves user efficiency. But the major benefits will accrue when the IT group doesn't have to wrestle with reintegrating software from different companies at each new release. It's no longer our problem; it's theirs.
But sometimes we weren't so sure we had made the right choice. After nine of our branches and our back office had been converted, we discovered our version of the software had scalability problems that prevented us from adding additional users. Going in we knew our installation would be the largest of this particular version, but we had accepted the assurances that the product would scale. When we met to discuss options, no one was smiling.
Fortunately, we never got to the finger-pointing stage. We were pleasantly surprised by Intuit's approach; delivery of the new version isn't canceled, only delayed. We moved forward with a previous version of the software that provided all the functionality of the new release but was character based rather than graphical.
Whenever things looked shaky, I reflected on the contract that kicked off our partnership. Did we negotiate what we needed? No matter how good the contract, recovering costs from a failed software implementation is nearly impossible. If it's true that the best contract is one you never have to look at, we did OK.
If We Had to Do It Over ...
But some developments fit under the How I Would Do Things Differently If I Had to Do It Over category. We ended up eight months behind schedule, and after we do the postmortem we'll probably be 20% to 30% over budget. While this isn't on par with some of the ERP horror stories you read about, there was room for improvement.
Defining standard operating procedures and naming conventions is an emotional event. People are funny about change. Talking about things that directly affect them without a neutral third party makes for drawn-out and costly events. Deciding how we were going to renumber our office, for example, took the better part of a day and a half.
Having someone with an egg timer, no opinions and authority would have been a blessing. And had we repeated the phrase "Does it really matter?" during the writing of 400 procedures, it might have saved us a bundle.
Further, we could have learned more had we not only rolled out the system in five phases but also paused in between. Deferring the final phases for as much as three months would have allowed us to polish procedures, discover lapses and involve associates from unconverted branches to participate. But the tyranny of the calendar, the complexities of crossing year-end boundaries and previous project delays didn't allow this luxury.
In our first phase, we converted our back-office functions. Less than 10% of our workforce makes up accounting and other support services, and most of their activities are seen only internally. During this conversion, we polished our training manuals and refined our PC requirements.
During our next two phases, representing another 10% of our workforce, we discovered weaknesses in our branch IT infrastructure and standard operating procedures. North Coast users, who could translate between the old and new ways during the final two phases, were trained. This was also when we uncovered the scalability issue of the new software.
Though we had learned much in the earlier phases, on the Monday morning following our next conversion, our trucks waited by the loading dock of our largest service center. The warehouse crew struggled with the new procedures. Our customer crews waited at their job sites for the deliveries.
With a little more training and a little more practice at sites that were up and running, our warehouse crew would have been less flustered. We didn't make the same mistakes during our final conversion.
Why It's All Worth It
Even with schedule slippages, cost overruns and midcourse corrections, I am still excited. Though we aren't off vacation from our disgruntled customer, we have a story to tell, and it's a better one than before. We've seen cost savings. We can accommodate new business opportunities.
Now I just need the ERP project to give me a vacation, far away and paid.
Les Johnson is CIO at North Coast Electric Co., a wholesale electrical distributor in Bellevue, Wash. Write to him at ERPJourney@ciodecisions.com.