|
Also See:
New Tools Finds, Rates SAP Consultants
ASUG's Masney On SAP's Direction, Customer Care
Mike Kerrigan, vice president of business applications at IT consulting firm Laurus Technologies, has helped more than 100 companies across all industry sectors deploy SAP business software. His years of experience have given him a unique perspective on SAP deployments—including why some projects succeed and why some fail. Here are his five steps for CIOs looking to ensure SAP success.
Mike Kerrigan is vice president of business applications at Laurus Technologies, an IT consultancy based in Itasca, Ill.
During his career, Kerrigan, a one-time chief operating officer, has helped more than 100 companies across all industry sectors deploy SAP business software. Among the companies he has worked with are Geico, Phillips, Rockwell, and Chevron
Back in 1994, he worked on one of the first mid-sized SAP implementations in the U.S.—he was responsible for implementing materials management and production planning including all core SAP modules and strategic corporate functionality using SAP's Profitability Analysis. The implementation methodology used on the project eventually became the SAP ASAP rapid deployment methodology.
His years of experience have given him a unique perspective of SAP deployments—including why some projects succeed and why some fail.
He recently spoke with CIOZone chief content officer John McCormick. This is an edited version of their conversation.
CIOZONE: Granted, just about any information technology projects worth doing will have its share of challenges. But why do we see SAP implementations fail?
KERRIGAN: To me, a lot of times it's leadership and not understanding what you're truly trying to do with the enterprise.
The reasons people don't get the return on investment they expect are mostly people oriented.
My biggest issue over all these years is politics. Get rid of the politics—not just in these global companies, but in small companies too. When everybody is on the same page, with some leadership, and agree on how they're going to define key performance indicators and how they're going to use the software to benefit the company—those are the successful efforts. Everybody else just keeps fighting.
Why do we still have failures? Because of the people. Because of the decision-making. It's because the organization is still siloed and is unwilling to accept the integration standards that you have to have.
CIOZONE: People say they know this, yet we keep seeing problems. Just think of the list of companies that have struggled with SAP—Goodyear, Waste Management, American LaFrance. And, in the case of American LaFrance—which emerged from Chapter 11 in May—the company blamed a botched SAP implementation as one of the reason it sought bankruptcy protection.
KERRIGAN: I can't speak specifically about those because I haven't been there.
There is another company, which I won't mention, that did go bankrupt. And I helped them all the way to the end with a trustee. It was probably one of the worst implementations I've seen. And it was done by one of the big five consulting companies.
What I mean by the worst implementation is I couldn't believe the things that they missed that we knew about as consultants in 1994. So, unfortunately, in these situations the companies rely on their consulting firms, and if the consulting talent only knows a few ways to do it—and can't think outside the box—there will often be a failure.
SAP is driven toward integration from the business process standpoint. But all of that is determined by your master data. And I think consultants don't explain that to companies well enough. If you look at companies that fail, you need to ask what their real problem was.
Were the processes broken? Were they trying to do something that no system can do?
Or it could be that they never understood how important master data is. Is your master data correct? Is it doing what it's supposed to be doing? Are you missing fields?
You know, if you look at old legacy systems, logical files and everything else, your master data was spread across many things. But in SAP a material master has well over 500 options and fields. You're taking legacy from here to there. At Laurus, we've actually built a managed service around master data, so we can help companies validate it.
CIOZONE: But what about the tools—data cleansing tools—that can go in and clean up your data.
KERRIGAN: There are cleansing tools, but many of the tools only do half the job. That's why ours is a managed service that has all of the tools combined. Ours is all about a methodology of going through this. And, as we go through it, the customer can interact constantly with the data to see if something that comes up as a duplicate is really a duplicate, to see if suspect data is really invalid. And we can make a decision to fix it before the data is put into the SAP system.
CIOZONE: What are the top five things a company should be doing if they want to ensure that their SAP implementation will go off without a hitch? Obviously you think master data should be on the list.
KERRIGAN: Number one—team integration and leadership. Leadership is something I just don't see anymore. This is a personal opinion, but I think I've seen America almost be like the Japanese and its decision by committee. When you get into decision by committee without leadership, things stall. Things don't get done.
When I first worked with SAP, we were successful because we had direct contact with the executives and we had the leadership. And we did workshops with executives to tell them what was going to happen, how it was going to happen, and how we needed them in the process. And we got their buy-in before we ever even loaded SAP. That's a key component.
And the teams need to be integrated. If all the team members agree—manufacturing agrees with sales and sales agrees with accounting— obviously it's going to go a lot smoother.
Second is understanding the entire driven force of master data. You have to get that right. There are tools and there are processes. I would tell a client today that if they're going to start an implementation of SAP to start the project by at least taking their legacy data and understanding it and cleaning it up. How many duplications do I have? What's missing? You know, SAP's going to add 200 fields to it, but at least I'm taking a subset of pre-cleaned data to start that process.
Another big issue is training. If we go back over time, the number one piece of the budget that always seems to get cut is training. But training has got to be done. Everybody you talk to out there—from a user standpoint—always goes, "god, I just didn't get enough training."
And where's the integrated training? What I mean by that is people in the warehouse need to be trained on what sales does, and people in sales need to know what the warehouse does. And people in production have to know what the warehouse does. And people in finance need to know what all three do.
And it doesn't take that long, but at least now people understand the inputs and outputs of why something might be broken, why it could be the material master.
Then there's security. Again, almost like master data, you've got to have a plan up front, especially with Sarbanes Oxley, or SOX, regulations. If a company is public and either has SAP or is thinking of implementing it, they'll probably tell you—under a deep breath—that security's killing them.
The fifth is the technical infrastructure. You really need to understand the new technology under SAP. You need to know that SAP is going to have Java engines underneath. And so you better get the right experienced people established to make sure those systems are up and running.
You have now virtualization in the databases. You have failover. You have all these things.
If you go back 10 years or so, we had one server for production, one for quality assurance, and one for development. Now people have 50 or 60 servers trying to run things, especially if you have any of the SAP cross applications—Customer Relationship Management, Advanced Planner and Optimizer, Supplier Relationship Management. The more that SAP grew their footprint, the more complex the technical end became.
CIOZONE: You're talking about a good deal of infrastructure expertise. But do companies have these skills resident or do they usually have to go out and hire some people?
KERRIGAN: They might have to hire some people. Or they could get consultants, who could then train the staff.
But, in most cases you're going to need more skills—Java experience, for example—and people don't understand that.
Companies really need to take a different view of bringing in younger talent and spending the money on training them.
Only registered users can write comments.
Please login or register. |
