inventory optimization

Replies to This Discussion

Permalink Reply by Ira Gershkoff on April 2, 2009 at 3:49pm

It's pretty hard to give a specific answer to your question, since it depends on the nature of the inventory you're tracking and optimizing. However, I can give you some general guidelines that I use.

If you use a packaged solution like SAP, there should be at least an 80% fit between what the package offers and what you need. You should always expect some customization of a package solution, and if it's more than 20% of the functionality you need, then in my experience it's not worth it.

Similarly, the cost of the package solution should be a lot lower as well. It should be no more than 25% of what you would spend on a fully customized solution. Then expect to spend another 25% on customization and integration. If the cost of the package software + your integration cost is on the order of 50% of the cost of a customized solution (or less), then using the package may be a viable proposition. Much beyond 50%, the risk of integrating the package with the customized integration is no longer a good tradeoff.

Ira Gershkoff
SlipStream Aviation Software
703-297-5818

▶ Reply to This

Permalink Reply by Yi-Chun Tsai on April 3, 2009 at 7:18am

Hi, Ira:
Thanks for your input. My company is a online retailer that sells MRO kind of products to mostly manufacturing sector. One article that I read said that if your industry is not covered by a package solution then customized solution should be considered, which probably is not my case.
Another question that I have is that if a packaged solution is the way to go, then does it mean that anyone can take up the job of optimizing inventory? I have backgroud in Operations Research and I guess my question is that what is the value of having that kind of background if anyone who can use a packaged solution can do inventory optimization! Thanks.

▶ Reply to This

Permalink Reply by Ira Gershkoff on April 3, 2009 at 7:45am

To DESIGN or BUILD an inventory optimization solution, you need your kind of background. To develop procedures for managing inventory, you need your kind of background. To OPERATE an inventory management system, you shouldn't need more than a high school education and some common sense. The system should hide as much complexity as possible, so if you need a PhD in Operations Research to manage an inventory control system, then it's not a very good system. The value of your background is that you can formulate, develop, and implement the system so that someone less expensive than you can operate it and realize its value, thereby freeing you to go do something else.

▶ Reply to This

Permalink Reply by Yi-Chun Tsai on April 3, 2009 at 8:23am

Hi, Ira:
Thanks for you input again. I do have a PhD in Operations Research even though my research is on how to assign so called "flexible servers" among different tasks dynamically so that either the whole system throughput can be maximized or the average operating cost can be minimized. I am not an expert on Inventory Optimization but I do have some familiarity with it and I know that it involves Mixed Interger Programming which is a hard problem to solve.

My problem is that how do I convince the people who are currently in charge of it that I can help in their work? I told them previously that I am interesed in helping them but no one is asking me to currently.

▶ Reply to This

Permalink Reply by Ira Gershkoff on April 4, 2009 at 9:08am

You've put your finger on the marketing challenge. In order to convince people that you can help them, you have to bond with them and show how your approach will help them. That means you need to understand enough about their business to show them that you can develop solutions that they can use. This is usually more difficult than developing an optimization model, but getting your clients to be comfortable with your approach is a key part of the product you're selling.

▶ Reply to This

Permalink Reply by Sharethram Hariharan (Shareth) on April 7, 2009 at 10:16am

Awesome and informative replies Ira. Thank you - Many discussion groups (not only in analytic Bridge) have this topic of either purchasng a solution like SAP/i2/ManH versus customized solutions vs what we can build in-house. Almost all of them seem to agree that assessing your own needs and clearly identifying its role in the long term is the key to your decision. Also solutions like SAP etc have a higher cost to upgrade and maintain and we also have to take that into consideration.

▶ Reply to This