Cost analysis for SQL Server

[Gwen Thomas <gthomas -at- PAYSYS -dot- COM>]

Usually I'm careful to only read and post to Techwr-l during lunch hours or after I'm off the client's clock, but today is an exception.  Several threads here concern exactly what they're paying me for, and I sure appreciate all of your contributions!

So here's a current (since lunch) situation.
The client (a software developer) has multiple databases holding similar kinds of customer-related and product-related information, as well as a set of documentation databases and a huge call center database. They're in everything from Access to Foxpro to 16-bit Sybase Anywhere to whatever. Right now these databases aren't related, but the intrinsic relationships exist (thanks, Mark, for the terminology), so the logical modeling is the easy part.

In theory MIS has decided to go to MS SQL Server as a company standard.  The company (which has personnel on 6 continents) is in the process of migrating from Novell Groupwise to MS Outlook and plans to standardize on MS products as much possible.

I've been working with various groups on how to migrate all the various information sets into a comprehensive set that all groups can access.  Everyone's in agreement that the whole will be MUCH more than the sum of the parts.  Several groups are planning to build Visual Basic, Access, or ASP front ends to this common back end or to run Word mailmerges and other basic office tasks off this common data store.

Timing of the migration will depend on resource allocation, other project timelines, and purchasing/leasing of client licenses.  I've been working with the #2 MIS person for a while now on this.

Now she tells me that the bean counters (justifiably) want to know why they should spend $50,000 on SQL Server client licenses when they can get a run-time version of Sybase SQL Anywhere for just a few thousand a year.  They're willing to spend the money, but they want a serious cost justification here.

I can and will provide an analysis of the benefits of planned projects.  But this doesn't address the Sybase vs. SQL Server issue.

Common wisdom is that they would encounter a slew of issues if they use the Sybase product.  But the bean counters don't want common wisdom.  They want facts if possible, or at least informed opinions.

So, can anyone out there help?  Any experiences in this area?  The way we see it, it doesn't take long to eat up $40,000 savings if you're constantly tweaking stuff or if key personnel working off the web can't get into data reliably.  At the level of usage anticipated, even small issues could add up to a large bottom line.

BTW, we're certainly willing to be convinced that SQL Server is not the solution.

Thanks in advance,

Gwen Thomas
Knowledge Management Consultant
CIBER Information Services


 From ??? -at- ??? Sun Jan 00 00:00:00 0000=