Loraine Lawson spoke with Jim Honerkamp, CIO of The Hillman Group, a supplier of fasteners, key cutting machines and engraving equipment to national, regional and local retailers.
Lawson: Can you give us an idea of how your IT system operated before this project?
Honerkamp: This company was a growth through acquisition proposition. Historically, we did not consolidate into a single system. Our master data really resided in three separate transactional systems and there was a lot of duplication and inconsistency. One is a home-grown transactional application called CCS. That supports our fastener product line. We also have JD Edwards World, which supports our letters, numbers and signs product lines and engraving. And we are moving to a new ERP system, which is Oracle Enterprise One.
A lot of our big box customers require us to deal with them using global data synchronization, and in order to do that, we had to publish our item attributes to a public data pool. The only way we could do that - because we had this inconsistency in our transactional systems - was to basically cleanse that data and house it in a master spreadsheet. We ended up with a fourth repository - the spreadsheets. We decided we had to come up with a product information management solution (PIM). So we came up with iSPOT - that stands for item Single Point of Truth. This was going to be the Bible for our item master file.
Lawson: How does it work now?
Honerkamp: Basically, iSPOT is a relational database for placement of all item (sku) attributes. All of our item maintenance is performed in iSPOT now. We don't do maintenance within the transactional systems to the item master file. Everything is done in iSPOT.
We use the Ascential tool, which is now the IBM Information Server Quality Stage tool. We bought the product right at the time IBM acquired Ascential, so all the presales were Ascential guys who became IBM guys. We use the Quality Stage tool for two purposes. [The first is] to synchronize our transactional systems with the updates that we do in iSPOT, so if somebody changes an attribute to a specific sku in iSPOT, we want to make sure that's changed in transactional systems, so the IBM tool does that. We've written scripts that pull the information out of iSPOT and recognize what needs to be updated and does that update. That's the back end. If you think of this as a loop, we load iSPOT from the transactional systems, and that's a one-time situation. In theory, it's a one-time deal; in practice, every time we acquire a company, we'd load that again using the the IBM Information Server tool. The second is we load iSPOT using the Quality Stage tool and on a nightly basis, the back end of that loop takes place, where we extract any changes made in iSPOT and pump those changes back into the transactional systems. And all that is done using the Quality Stage tool. And any change to any sku attribute is done in iSPOT.
Our big box retailers require us to publish some data that's not native to our transactional databases - like marketing information, for instance. Home Depot, for instance, has images and marketing information they want to appear on their Web site. We don't physically store that information in our transactional systems. So iSPOT isn't only a solution for cleansing data and storing core attributes; it allows us to extend our item information with things like marketing data and images and not have to go back in and modify our transactional systems. There's an attribute of iSPOT that prevents us from having to go back in and do a lot of maintenance on our transactional systems.
Lawson: What were your goals for the project?
Honerkamp: The project goal was to meet our customers' requirements. Lowe's, Home Depot, Wal-Mart - all the big box retailers are forcing suppliers like us to participate in global data synchronization (GDS) and we knew it'd quickly get out of hand if we tried to do this via spreadsheets. We also wanted to do an online product catalog so we could do business via the Internet. We wanted a single point of truth for our products catalog. So that's where we came up with the idea of a Product Information Management solution using the Quality Stage tool as a data pump into and out of that single point of truth.
Lawson: What were the challenges in terms of choosing a solution to help you?
Honerkamp: Affordability was one. The overall design was very challenging. That's more along the lines of iSPOT itself. The Ascential tool, the Quality Stage tool, was pretty straightforward. When I was first exposed to the tool, I knew right away that was the tool we needed to address this requirement. We didn't really see anything else on the market that was quite as robust as that tool. We did look at Information Builders - they had an ETL tool we were familiar with. That was not nearly as robust as the IBM solution. In picking the tool, it was fairly straightforward. The design of iSPOT itself was quite a challenge.
Lawson: There wasn't a solution that could do everything?
Honerkamp: IBM had a solution, but we couldn't afford it. We're a half a billion dollar company; we've got limited resources. They had an excellent tool. Had we been able to come up with the capital for it, we would've bought that solution. We created what I call a poor man's Product Information Management solution. We designed the repository on our own and then wrapped it with the Quality Stage tool. The challenge there was we needed a solution that was item-number centric and GTIN-centric - a global data synchronization term. We basically used a tool from Lansa - an IBM partner - to help us design iSPOT. We've owned it for years.
Lawson: Any idea on when you'll achieve an ROI?
Honerkamp: We didn't justify the project or the purchase of the tool on ROI. It was really more of a cost of doing business. We have to provide this service for our customers, so how can we do it in the most efficient manner? We justified the purchase of the tool really across two projects: the iSPOT project and we also recently implemented an advanced planning and scheduling application that requires us to move massive amounts of data from the transactional systems into our APS systems on a daily basis. We're moving over 6 million records and we've got to get it done in a 15-minute window nightly. The Quality Stage tool is the only tool that can do that in that time.
Lawson: Any lessons learned you'd like to share with other CIOs who may be facing such a major integration challenge?
Honerkamp: The challenge is global data synchronization and how companies are going to efficiently and effectively meet the demands put on them by the big box retailers, although second-tier retailers are now getting into this as well. How do you meet the demands? How do you remain flexible enough to meet ever-increasing demands of speed and accuracy with big retailers? The creation of a PMI solution is absolutely necessary. If you try to do it in spreadsheets or any other way, if you have multiple transactional systems - and more companies do than don't - you're setting yourself up for failure. We realized early on we had to have a single point of truth and then the challenge became how do we do it within the budget we have and so it's maintainable by our people and how do we make sure our back-end systems keep up. If we didn't have an automated way of doing it, at least for us, we were looking at a whole lot of duplicate work on an ongoing basis. If you can buy a solution off the shelf, great - if you can afford it. They're not cheap. And, an off-the-shelf solution doesn't address the synchronization back to your transactional systems - none of the ones we looked at did. So that's where the Quality Stage tool came into play for us. That was the glue to tie it all together.