Cisco Case Analysis

EXECUTIVE SUMMARY

January 1994 Cisco's legacy system surpassed its standard modification requirements and encountered a malfunction that resulted in corrupting the database. The company was almost completely shut down for two days. It became clear that the legacy system would not continue much longer and a solution was required.

Cisco was faced with the dilemma of selecting from three potential solutions to the system issue: 1) Upgrade to the new version of the legacy, 2) Implement a single integrated replacement of all applications in parts, 3) Implement a single integrated replacement of all applications as a whole. After careful analysis of the three alternatives I chose to support the third recommendation. Details defining this issue and supporting my recommendation can be found within the remainder of this document.

Cisco Systems, Inc. a technology-based company was founded in 1984. By 1993 Cisco had grown to a $500 million company. However, Pete Solvik the company's CIO projected growth to $5-billion plus per year in the future. The UNIX-based software package that had been used to support all core business processing could not sustain the current level and certainly not future growth.

By January 1994 the legacy system surpassed its standard modification requirements and encountered a malfunction that resulted in corrupting the database. The company was almost completely shut down for two days. It became clear that the legacy system would not continue much longer and a solution was required. The question that demanded attention within this case pertained to how the legacy system should be replaced.

The current system had coincided with Solvik's organizational and budgetary approach that allowed budgetary decisions to be made at the functional level, but left IT organizations reporting directly to him. This allowed the modifications required by the legacy system to be addressed by each functional area as required. Ultimately, these modifications ("workarounds") contributed to the systems malfunction. Ironically, the three alternatives produced to address the issue were all consistent with initially removing the functional level budgetary responsibility. This was based on two reasons: 1) the lead-time involved in the implementation would not allow for multiple areas performing their own investigation, and 2) the functional areas could not justify the investment of at least $5 million each.

The factual content of the case provides Cisco's system requirements and current limitations, along with budgetary considerations. First of all, Cisco was in a position that they required a replacement system immediately. The current system failure had proven that it could not be ignored and could ultimately cost the company by losing data completely, as well as potentially losing revenue. The current software would not support redundancy, reliability, or maintainability. It had become too customized and could no longer support the needs of the growing business. Without this functionality, Cisco would be risking streamlined customer support. This risk could not be undertaken with the 80% annual growth Cisco had been enjoying.

Three alternatives were produced that would be instantiated immediately upon approval. The three alternatives consisted of: 1) Upgrade to the new version of the legacy, 2) Implement a single integrated replacement of all applications in parts, 3) Implement a single integrated replacement of all applications as a whole.

The first alternative considered upgrading the current legacy system to the latest version. This alternative has a couple advantages. First of all, Cisco is the largest customer of the current software vendor; this capacity could be used as leverage. Secondly, it would be more cost effective to initiate an upgrade versus a full-blown replacement.

However, there were disadvantages as well. The current vendor was undergoing new ownership. The unknowns associated with new ownership left questions as to continued support. Would the new owners of the legacy system provide the level of support required for this project? Could cost benefits still be achieved based on Cisco's capacity? Both of these questions were risks involved. Even more importantly, the legacy system with upgrades would still not fully support Cisco's projected growth potential. If they were going address the issue it should be addressed for the long-term and not just another "band aid".

The second alternative considered replacing the legacy system in parts. Implementing the replacement in an incremental fashion did offer some advantages. First of all, the effect on the IT department could be controlled. Each increment could be approached as simply another project within the "normal" business functionality. Secondly, there would be transition time between the functional areas of the company. Issues could be uncovered and dealt with as encountered in lieu of affecting the entire operation at once.

However, this approach did have disadvantages as well. The incremental process could take on a life of it's own by allowing the functional areas to customize too much and lose sight of the streamlined necessity for future growth needs. Ultimately, this approach could become enormous and drag on for some time. Secondly, there would be risk in maintaining various systems while the replacement occurred in parts. It would be unknown if various systems would work together and this could lead to limited functionality between the functional areas. Lastly, replacing the system in parts does not achieve "buy-in"