Monday, November 15, 2004

Is Excel the poor man's i2?

Enrico Camerinelli from Meta Group wrote me this week with a hypothesis that's not going to sit well with the supply chain planning (SCP) vendors.

In the market for SCP software, it could be argued that the market leader is not i2, Manugistics, or SAP, but Microsoft Excel. Many companies never look at the leading SCP vendors, finding that it's much simpler, easier, and certainly less expensive to simply create an Excel model to plan short term production around one or two key constraints.

Enrico says that the advantages of using Excel are many:
  • Overall ease of use.


  • Allows the planner to use his knowledge and experience to create a model that is highly relevant and specific to the problem.


  • Allows the planner to quickly generate multiple scenarios.


  • Allows the planner to tweak the results if needed.


  • Allows the plan to be annotated to explain the results or provide clarification.

The downside, of course, is that an Excel model for supply chain planning is likely to be disconnected from backoffice systems, creating extra work to download or rekey data to the model and to import the results of the approved plan back into the production system.

So, what do you think? Do you think that using Excel is a viable alternative to buying a costly SCP system? Click on the comments link below and give your view. Or send me an e-mail.

Update, Nov. 19: Karthik Mani, a VP at i2, saw this post and wrote his own post regarding the advantages and disadvantages of Excel for supply chain planning. He writes,
Excel has its advantages. Some of them are
  1. Familiar interface. Planners use it every day.

  2. Desktop availability very similar to web, but planners can display a lot more data than can be provided in a standard web based tool.

  3. Very easy configurability for the end user - layout, pivots, filtering, graphs, logical functions, math functions.

  4. Very easy availability of skilled resources to help end users with more sophisticated changes to the engine.
But Excel based planning also has its big disadvantages
  1. In even medium sized enterprises, multiple people are responsible for coming up with the operational plans (demand plan, staffing plans, production plans, logistics plans etc.) and Excel based planning is very poor at synchronizing those plans.

  2. Excel models will be disconnected from the back office systems leading to painful keying in of master data.

  3. Enterprise has no control over assumptions and logic used in the planning process.

  4. Communication of metrics (this quarter we need to focus on revenues rather than market share and profitability, next quarter we need to focus on market share rather than profitability and revenue) is very tough.
Karthik thinks that the ideal approach is to marry the use of Excel with a best-in-class SCP solution. As an executive at i2, it's not surprising that this would be his view. But I think it does make a lot of sense. He continues,
The best in class SCM solutions have now moved to marry the advantages of both of these approaches. Here you have a central data and plan synchronization infrastructure. The front end exposed to the users is Microsoft Excel. Some of the light planning logic can be directly incorporated into the Excel sheet; all of the reasonably sophisticated calculations will reside in engines that will be available to the spreadsheets as web services.
Read Karthik's post on his weblog.

2 comments:

Frank Scavo said...

Dwight Shih e-mailed me with similar observations regarding financial analysis applications:

If you spend some time with Financial Analysis Software vendors, then I think that you'll find their toughest competition is existing excel spreadsheets in the hands of financial analysts. The counter argument is often not the accuracy of the spreadsheet, but rather the transparency, consistency and provenance of the spreadsheet:

1. Who developed the spreadsheet?
2. What assumptions are built into the spreadsheet?
3. Are the formulae used to evaluate B the same as those used by a colleague to evalulate A last week?
4. Did the analyst [intentionally or otherwise] tweak one of the underlying assumptions for this analysis?

I suspect that there is a market for an excel specific version control system. One that could open up the spreadsheet and highlight both formulaic and data differences. Then analysis could be tied to a specific version of a specific spreadsheet variant.

Posted at http://www.ideoplex.com/blog/2004/11/17.html#a1034

Karthik Mani said...

Here are some further thoughts on this topic in my blog.

http://www.bloglines.com/blog/KarthikMani?id=13

Other areas beyond SCM seem to be thinking the same way. Dwight had alluded to this one in his comment.

Karthik Mani