What is our primary use case?
I was part of a standards organization and we built a data model that is a standard data model for use in retail. That data model is now been released in version 7.3 and it is implemented all over the world. We don't implement the model, we've built the logical model and then the companies build their own physical model from there.
erwin is a retail data model, which means that it handles the operational side of retail, which means there are somewhere around 8,000 attributes in it. It has got around 10 groupings of things. We have a grouping on transactions and there are all kinds of transactions that can occur in retail. The whole customer life cycle is covered in the inventory, items, and all that. The use case is for retail operations. It's massive. There are hundreds of use cases in this.
How has it helped my organization?
We don't implement, we simply tell other people how to do it. It's a beautiful model for the new microservices world, so we can help people understand how to fit this into their world. In terms of us actually doing something and implementing it and all that, that's really not in scope for what we do.
erwin is easy. In the microservices world, having a unified retail model like this one that is a standard and allows two companies to inter-operate easily in the past. In fact, the whole reason the model was created was in 1993, was because about half a dozen major retail CIOs got together and said, "We've got to have a standard model because every time we buy a new point of sale system, we need to re-architect our entire enterprise." They started building this model back in 1993, and the beauty of it is it does precisely what they say. A retailer can now integrate two vendor's systems easily, as long as they all follow the same model. It reduces their cost of integration dramatically, as well as being quite a powerful model in and of itself.
It reduces monthly savings by hundreds of thousands of dollars. Think about a company like Costco and all of the points of sale systems in Costco, all of the systems, and the applications, but if all the applications in Costco all had their own data model, trying to integrate those, upgrade them and manage their different versions of the same model throughout the store, is an absolute nightmare. It's phenomenally expensive. This helps reduce that cost significantly. I'm talking on the orders of hundreds of thousands of dollars.
What is most valuable?
erwin is pretty easy. I've been using it for so long it's like second nature.
The visual data models are pretty easy for helping to overcome data source complexity and enabling understanding and collaboration around maintenance and usage. It's easy to add, change, and update things. We get feedback from retailers. For example, somebody wants to update something in the item area, they want to use a new item identifier and it's just a matter of going in and adding it to the numerations for that. Or somebody might come in and say, "We're using a little bit of a different pricing model so we need to add this information into the pricing area." Or people will say "We need to add Bitcoin," so we can go in and add Bitcoin and the attributes you need to support it and do it very easily. At this point, we're not adding new capabilities, we're simply expanding existing ones.
What needs improvement?
The navigation is a little bit of a challenge. It's painful. For example, if you've got a view open and you want to try to move from side to side, the standard today is being able to drag and drop left and right. You can't really do that in the model. Moving around the model is painful because it doesn't follow the Windows model today.
Otherwise, it's got everything I need and it's not hard to use for me.
What do I think about the stability of the solution?
The stability is great. We don't have any problems.
How are customer service and technical support?
I actually did use their support, I had some issues getting it installed and it had to do with that they've given a copy of the Data Modeler for me to support the standard data model, and getting that approved and authorized and all that was a bit of a challenge. I went through the help desk and they got it done pretty easy for me. I had a unique problem.
Which solution did I use previously and why did I switch?
I had used D-Base. This was a long time ago, but I used D-Base to build a model for the oil industry. That was a long time ago. It was a 1980s vintage so there is no comparison.
How was the initial setup?
The initial setup is straightforward. You can install it without a lot of hassle.
What's my experience with pricing, setup cost, and licensing?
They gave us a copy because of supporting a standards data model, so pricing and all that is really not something I can compare. I think it's a bit expensive, but it supports and does what we want.
Which other solutions did I evaluate?
At one point we had a data modeler who wanted to switch to Embarcadero, and it turned out that that was a huge mess so we dropped it. It didn't last very long, but it was a data modeler who came in and wanted to do it in Embarcadero. I think she had an agreement with them and got a bonus for trying to get it converted or approved to convert but it was such a huge mess we didn't do it.
The Embarcadero model is huge. It's got 8,000 attributes in it. Being able to go through and validate that every one of those 8,000 attributes properly converted over to the correct place in Embarcadero was such a massive job. We didn't mess with it. It's not just the attributes, but it's the relationships and table names. It was a huge job so we didn't do it. I suspect if we had gone to Embarcadero, it would have been just fine, but it was just too big of a job.
What other advice do I have?
erwin DM is good. It does the job and it's been around a long time, so I think it would be a good one to use. I don't have any problems with it.
I would rate erwin DM a nine out of ten. Nothing is perfect. I don't have any real issues with it. It does everything we need it to do. It's really good.