home *** CD-ROM | disk | FTP | other *** search
- Comments: Gated by NETNEWS@AUVM.AMERICAN.EDU
- Path: sparky!uunet!stanford.edu!bcm!convex!darwin.sura.net!paladin.american.edu!auvm!LOBBY.TI.COM!RSNYDER
- X-Mailer: ELM [version 2.2 PL16]
- Message-ID: <9211131827.AA06716@lobby.ti.com>
- Newsgroups: bit.listserv.sas-l
- Date: Fri, 13 Nov 1992 12:27:06 CST
- Reply-To: "R. Snyder" <rsnyder@LOBBY.TI.COM>
- Sender: "SAS(r) Discussion" <SAS-L@UGA.BITNET>
- From: "R. Snyder" <rsnyder@LOBBY.TI.COM>
- Subject: Re: SAS Consulting Services
- Comments: To: 71020.1025@COMPUSERVE.COM
- Comments: cc: sas-l@uga.cc.uga.edu
- In-Reply-To: <9211130312.AA03811@lobby.ti.com>; from "William Kahn" at Nov 12,
- 92 9:18 pm
- Lines: 61
-
- SUBJECT: SAS Consulting tips.
- SUMMARY: Get your ducks in a row.
- E-ADDR: rsnyder@lobby.ti.com
- NAME: Bob Snyder
- COMPANY: Texas Instruments, Sherman, TX
- PHONE: (903) 868-5799
- FAX: (903) 868-7240
- >
- > Subject: SAS Consulting tips?
- > Message-ID: <921113021834_71020.1025_EHC28-2@CompuServe.COM>
- >
- > We are just starting a small project with SAS Consulting Inc. (a subsidary
- > of SI). <...> Has anyone on the list any experience working with SC?
- > Any suggestions on what we can do to maximize the chance of a successful
- > project?
-
- > <...> The "look" of the graphs is very hard to define contractually,
- > but very important in terms of value.
- > The underlying data structures used favor certain kinds of future
- > expansion over others. Etc.
-
- > Does SC make "good" choices? How "micro" should we manage this project?
- > Are there an important differences in which programmer actually does the
- > work? Do you have specific recommendations?
- >
- > William Kahn
- > W. L. Gore and Associates
- > 71020.1025@compuserve.com
- >
- William,
-
- I personally worked with Don Henderson of SAS Consulting Services in Rockville,
- MA a couple of years ago to kick off a project here at TI. His help was
- invaluable and I recommend this service highly.
-
- The most valuable advice I can give is to get your ducks in a row before you
- start shelling out $800 + a day. No consultant can "make good choises" unless
- you know what you are trying to build. Specifically:
-
- a) Know who your users ( customers ) are, and what their objectives and
- expectations are up front. This will allow you to establish their data
- needs and hopfully their output reguirements ( reports and graphs ).
- Note: If a graph carries the proper information, it's "look" is fairly
- easy to tailor or change later on.
-
- b) Establish your strategic directions so that you can build the right data
- base. That is not easy to change later on.
-
- c) While I don't recommend micro management, I do recommend putting one of
- your own programmers ( train one if you must) on the team so that that
- the knowledge base SC brings to the project will to some extent stay
- with the project over the long term.
-
- d) The "which programmer" question is valid. Don was a specialist at building
- applications but was not a statistical consultant. Don't go off writing
- code until you know what kind of help you really need.
-
- I hope I've helped.
-
- Regards,
- Bob
-