Print

Print


Hi Jennifer & all

Rebecca Mander passed this email on to me, and I thought I'd let you 
know where we're at and also ask the group a related question.

I joined Brookes in 2009 with the task of developing a data mart plus 
university wide reporting tool to support our MI for the strategic plan 
(a subset of our KPIs around students). I started small with a a static 
prototype based on one source only. I would recommend the prototype 
route because it allows you to take the time needed to define the fields 
properly, sorting out user requirements and getting a complete and 
robust set of requirements and ETL definitions before you invest in 
external expertise. The prototype was delivered in Nov 2009 and we're 
now in the middle of an ITT to find a supplier who can
a) turn this into a 'live' data warehouse with data validation, 
automated refresh, slowly changing dimensions etc., using industry 
recognised ETL tools like BODI, OWB
b) provide a universally acceptable (!!!) reporting tool, again we've 
played it safe, major players (OBIEE, SAP BO, Cognos) only

We are also considering an interim stage: buying just the reporting 
software, and polishing up the prototype so that it's at least auto 
refreshed. ie we would omit the super duper time trend data warehousing 
development, and just have something up and running. Bearable in the 
short term and much much cheaper. It would not be futureproof in the 
long term as we bring in lots systems, but should be a cheap way to 
start bringing some of the core systems together in the first year.

So. First - my attempts to answer your questions
a) What products do you use? --- see above, procurement not yet completed

b) What solutions have proven most cost effective? - well a staged 
approach starting small has worked for us (sort of). Every other person 
I've talked to who has got the full implementation to work from scratch 
has spent £100-250K up front + lots ongoing. I do see that your top 
priority is to bring several systems together (this is our long term 
priority too) - but I would start getting the thing working just from 
one system. Also, using a contractor for us was much better value for 
money than a consultancy - but maybe we were just lucky.
In the long term, to have a robust workable system from many sources, 
you should definitely go the full data warehousing route, but you can 
get quite far in selling in the idea with something less formal.

c) Have there been many non-technical issues to resolve? -- some 
absolutely key things to get in place, which I have picked up over the 
years (iused to be a data mart project manager in the private sector):
--- commitment at the highest level, an evangelistic VC sorts the whole 
project
--- internal technical resource - 2 seems to be the minimum although 
possibly can work with 1 FTE, ongoing
--- internal business resource - a BI reporting team, strong user 
representation
--- this one's important - get all the requirements &definitions really 
pinned down to technical detail, as far as you can - this is where many 
projects fall over because they usually need the busiest people in the 
organisation to agree them, and no amount of time spent by IT or 
external people can replace this.
--- I know lots of people have had issues with reporting tool licensing 
structures, may not be relevant for you.

d) What would you do differently next time?
Um. Demand £250K up front? alternatively write a persuasive business 
case to ensure sufficient in house IT resourcing for the project.

*****************Which leads me to my question to the 
group:************************
-- I need to write a persuasive business case justifying additional 
resource in a difficult climate, by demonstrating the direct financial 
benefits of a Business Intelligence system. Any of you people with 
existing well established systems have any thoughts and/or examples on 
their benefits you could share, at the highest level? Anything at all 
would be appreciated, however vague. and anonymised of course.

Thanks - hope the above helps you as well

Georgina
-- 


Georgina Spary
Business Intelligence Manager
SBPO, Corporate Affairs
Oxford Brookes University
+ 1865 484887


-------- Original Message --------
Subject: [Fwd: Re: Reporting from multiple systems]
Date: Mon, 25 Jan 2010 17:49:21 +0000
From: Rebecca Mander <[log in to unmask]>
Organisation: Oxford Brookes University
To: Spary Georgina <[log in to unmask]>



-------- Original Message --------
Subject: Re: Reporting from multiple systems
Date: Thu, 21 Jan 2010 12:53:29 +0000
From: Andrea Cheshire <[log in to unmask]>
Reply-To: Academic, financial or space planning in UK universities 
<[log in to unmask]>
To: [log in to unmask]
References: 
<[log in to unmask]>



Dear Jennifer



We have invested in a data warehouse at UWE and have operated the system 
over the last 2 years. Currently the warehouse holds data from our 
student record and finance systems but we are expanding to bring in 
admissions data. The warehouse has opened up completely new avenues for 
business analyses which single systems can never provide . Although it 
is quite an investment, it is paying off in terms of efficiency and 
effectiveness of analysis and reporting. We are currently using it to 
drive performance improvements across the institution.



As the reporting front end we have chosen Business Objects, although 
there is very little difference to other systems like Cognos.



The key factor of success of our project was the quality of the data 
warehouse design. You really need to get this right, as it determines 
the avenues for future analysis. We used consultants as we don’t have 
the knowledge in house. However we ensured significant know how transfer 
in the first project and are now confident to do more ourselves.



The biggest non technical issue is the review of the quality of your 
data. We have spent months cleaning our records and are still picking up 
issues. But this is to be expected and I have had similar feedback from 
other institutions that went down the warehouse route.



We are happy to give more in depths informaton if this is helpful.



Best wishes

Andrea



Andrea Cheshire

Director Business Intelligence and Planning

University of the West of England Bristol

[log in to unmask]







*From:* Academic, financial or space planning in UK universities 
[mailto:[log in to unmask]] *On Behalf Of *Arkell J.H.
*Sent:* 21 January 2010 12:10
*To:* [log in to unmask]
*Subject:* Reporting from multiple systems





Dear all,



Here at Southampton, we're looking to improve our reporting from 
multiple corporate systems (rather than the single system reporting 
which makes up the majority of information at the moment).



I would welcome experience and advice from institutions who have been 
down this route already on issues such as



What products do you use?

What solutions have proven most cost effective?
Have there been many non-technical issues to resolve?

What would you do differently next time?



We don't have a data warehouse system at the moment, and I would be 
particularly interested in institutions who have found interim solutions 
to this problem without investing in a data warehouse.



Jennifer



________________

Jennifer Arkell

Corporate Planning Manager

University of Southampton

www.soton.ac.uk



023 80594447

(internal 24447)