In the early 1990s Walgreens hatched an
ambitious plan with the help of Anderson Consulting to replace their modest
TPF system with a layered Client-Server architecture based on Open Systems
standards. Although they ultimately managed to achieve their aim it was a
hugely costly enterprise which had to be completed (years) late by Unix
specialists. This was a TPF system that was less highly used
than the Sabre test system......
(from TPF Scoop circa: 1996/7 :)
Second thoughts at Walgreens ?
It's been two years or more since Walgreens began their project to
migrate from a medium-sized TPF system to a Client/Server configuration to serve their
network of over 2000 pharmacies across the continental US. Employing Chicago-based
Andersen Consulting from the outset, the Illinois-based Walgreens has found out for itself
how difficult it can be to replace an old, but highly specialized and efficient,
transaction processing technology, with a 'grab-bag' of the latest Client/Server
Unix-based systems.
Over the last year the roll-out of the new system has been halted
frequently, due to poor performance and complaints from the field. Costs have exceeded the
original budget considerably but Walgreens is now in the unenviable position of choosing
to either continue on a course which has no certainty of success or returning to TPF and
trying to salvage what they can from one of the most expensive R&D 'experiments' on
record.
All the classic, and by now well-documented, problems with Client/Server
implementations have surfaced so far at Walgreens. They have multiple layers of
independent products, which do not always work together, despite claiming to be 'Open';
they have had to extend a building to accommodate the extra Help-Desk staff necessary to
advise the field on using the new system; they have seen response times plummet for even
basic transactions that need to access any centralized information.
Perhaps the saddest part of this cautionary tale is that a Walgreens
publication described how the new system was only going to provide a portion of the
projected improvements for the entire project. The largest part was to come from Business
Process Re-engineering. If that piece had been successfully implemented, along with some
of the latest TPF advances, we could well have been reporting on a major success story in
Chicago. Instead we have the latest disappointment with Client/Server technology and the
visions of Andersen Consulting.....
