Building a Web-Based Reporting System: Lessons Learned

1
Building a Web-Based Reporting System Lessons
Learned

• AIR 2005

2
Presenters

• Jacque Frost
• Chris Maxwell

3
Presentation Overview

• Project plan and required resources
• Tool demonstration
• Project timeline
• Issued faced and lessons learned
• Working with Information Technology
• Working with the data owners
• Data issues
• Policy/Security issues
• Summery

4
Why a Web-Based Reporting System

• Meet increasing unit level data requests
• Provide data to non-experts
• Decrease OIR ad-hoc reporting volume
• Provide data standardization
• Provide roll-up to official totals
• Consistency in unit Strategic Plans

5
Planned Project Scope

• Web tools
• 100 thin-client
• Intuitive in use, zero training
• Reporting focus
• Most requested type of information
• Start with drilling into the Data Digest totals
• Will not be comprehensive or a full BI solution

6
Planned Resources Required

• Institutional Research (OIR)
• Will be responsible for application development
• About 0.4FTE for up to a year to establish a
  series of web tools
• Information Technology (IT)
• Purchase, install, configure software
• Set up and maintain server
• Troubleshoot as needed

7
A Typical Tool

• Student Enrollment Tool Restricted Access

8
Project Timeline

• November 2002 Development begins
• February 2003 First application published
• August 2003 First production tool published
• October 2003 First three Data Digest tools
  published

9
Project Timeline (cont.)

• February 2004 Salary tools and portal published
  under direction of provost
• July 2004 Sponsored Program Services (SPS) tools
  published
• November 2003 - Present More Data Digest and
  other tools are published, tool edits continue.
  Service also used for data collection and surveys

10
Issues Faced and Lessons learned
11
Working with Information Technology (IT) Issues
Faced

• Initial Resistance
• Did not see need for pre-BI solution
• Did not want to support non-IT chosen technology
• Not in line with ITs OnePurdue vision
• Agreement to OIR plan for project
• Additional support IDN datamart project
• If you cant beat them, join them!

12
Working with IT Lessons Learned

• Do not give up or give in on key issues
• If necessary, use the leverage of the OIR office
• Always look forward
• Concentrate on successes
• Leave conflicts behind

13
Working with the Data Owners Issues Faced

• Publishing web tools can make OIR look like the
  data owner
• OIR does not own the data - yet is providing it
  at any imaginable level of breakdown, at any time
  with the click of a mouse
• Encountered initial data owner resistance/concerns
  
• Data security/policy issues
• Loss of control over the information

14
Working with the Data Owners Lessons Learned

• Involve them very early in the process - show
  them the earliest drafts of the web tool
• Communicate continually
• Quickly address concerns and suggestions
• Use a sign off policy on tool release

15
Working with the Data Issues Faced

• Web tools are very demanding of the data
• Great data for campus reporting can look much
  worse on drill downs
• Data availability was a critical limiting factor
  in this project
• Time needed to build a typical new tool days
• Time needed to get signed off data months

16
Working with the Data Lessons Learned

• Do not wait until the data are perfect
• Release tools with notes and caveats
• In general, do not fix the data with application
  programming
• The web tool needs to agree with SQL database
  queries that use the same criteria
• Allow the web tool to expose some deficiencies,
  it can spur data owners and IT to make fixes

17
Security and Policy issues

• FERPA
• Drill downs can easily expose unit1 records
• Internal Purdue policies
• Might not be well documented in one case we
  received a waiver
• Goal is to provide access to anyone with a
  business need for the information
• No real sensitive information available with
  tools

18
Summary
19
Overall Results

• Project was a success!
• Proceeded on schedule with anticipated resources
• ITs IDN datamart project was a plus
• Tool use
• 3,000 primary queries by about 300 users in our
  first full year (Oct 03 Sep 04)
• Increasing traffic now averaging about 18
  primary queries per business day

20
Overall Results (cont.)

• Still the source for Purdue data over the web
• ITs BI tool still not in production
• Standardization of reported data a big plus

21
Overall Lessons Learned

• IT support role / OIR application developer role
  worked very well
• Used strengths of each area
• OIR control of tools enabled very fast turn
  around times for any needed edits / enhancements
• Data Digest totals validated the tools
• In retrospect, could have more aggressively
  marketed the project

22
Overall Lessons Learned (cont.)

• Tool logging is invaluable
• Technical issues were not the most challenging
  aspects of this project
• Data, people, and politics
• Accept politics as an integral part of the project

23

• This is exactly what I need. Now I can sleep
  peacefully over the weekend!

24
QUESTIONS/COMMENTS

• Office of Institutional Researchwww.purdue.edu/oi
  r