objectives
The top three objectives were Security (0.30), Technical (0.25) and Economics (0.25). Management received the lowest score of 0.20. Our first round of analysis determined that Security had the most influence in our Level 1 comparisons, while Management of the new software structure was assumed to be the least demanding of all four levels of our comparisons. The tie between Technical and Economics created questions about the first round of analysis. Since Technical and Economics were part of the company’s top four objectives, it was important to understand why they were considered equally important; the Technical and Economic objectives are completely different aspects of the business. Understanding the similarities and differences leads us to the eight criteria that are a level below the objectives.
7.1.2.The top three criteria
Protection is a criterion of Security and has the highest score of 0.17. Scalability is a criterion of the Technical objective and has the score of 0.15. Service Charge is a criterion of the Economic objective and also has a score of 0.15. Migration Cost was also a criterion of the Economic objective and has a score of 0.09. The tie between Scalability and Service Charge also led us to believe something was skewing our results. We began to ask ourselves whether we were having the right people to do the correct pairwise comparisons. At this point, we considered changing the individuals who were to perform pairwise comparison for each level of the model. The strategy ranking will shed more light to our decision.
7.1.3.The strategy rankings
IaaS scored 0.35, SaaS scored 0.34 and PaaS scored 0.31. Since we had a tie in the objectives between Technical and Economic, and a tie in the criterion between Scalability and Service Charge, it resulted in no true winner of the strategy. We knew then that we had to make some adjustments to our analysis of the model. This situation solidified our decision to create a second round of analysis.
7.2Second round of analysis
7.2.1.The top three objectives
Removing the cloud expert opinions from the comparisons that should have been made by the company did not affect the top three objectives rankings. Since Security decreased by two points, it showed that the cloud experts valued Security more than BMF’s upper management. This was good information to know, in that BMF should research more into the importance of security to ensure that all staff members buy into the company’s objectives, which is very important.
7.2.2.The top three criteria
Removing the cloud expert opinions from the comparisons reordered the criterion ranking and showed that Scalability was the most important objective to BMF, as it wanted to be able to grow quickly, with the lowest variable cost. Although Security is still important to BMF because it was one of its top three objectives, it is worth asking why Security should be the first objective when Protection is the third criterion? Migration compliance would be the answer to that question since it scored 0.13, which ranked this criterion in fourth place, with part of it beneath Security.
7.2.3.The strategy rankings
The strategy rankings did not change order, though SaaS lost two points and IaaS gained one, thus making IaaS a more definitive leader. With the second round of analysis and the new scoring of the model, IaaS is now the best choice because it scored significantly higher than SaaS and PaaS on Scalability and Protection. IaaS did not score the highest on Service Charge, meaning that it was not the cheapest. However, results show that it has the highest security rating. Therefore, BMF must spend more money on Service Charge to increase their customer’s data security.
8.Conclusion
BMF LLC is a small online fashion retailer that is in the process of scaling up its operations. To scale efficiently and maintain security, BMF must make the choice of which cloud-based strategy will work best; this decision is very complex. An HDM model was constructed and used as an aid to make this complex decision. It was built with four levels—Mission, Criteria, Subcriteria and Strategy. At first, through experts’ evaluation the HDM model evaluated all four levels. When the first model produced inconclusive results, the team re-evaluated and delivered a new HDM model delivery strategy. With this new delivery strategy, professionals from BMF evaluated the first three levels of the HDM model (Mission, Criteria and Subcriteria), while the cloud professionals evaluated the last level (Strategy). The new HDM model delivery strategy rectified the inconclusive results of the first model delivery, and identified IaaS as clearly the best choice for BMF.
9.Limitations and Future Work
The cloud service provider model can be used for any company planning on migrating their software services to the cloud. The senior management in such a company should perform pairwise comparisons for both the objectives and criteria and leave the experts to choose the best strategy. Experts’ opinions are also relevant for any other uses of the model. Here, the weights for the objective and criteria are relevant only to BMF because they are unique to its needs. The strategy weights are global and can be used for any other applications.
Future work should be conducted to find out which IaaS platform should be purchased. If an HDM model was created for this purpose, the pairwise comparisons should be conducted by the senior management of BMF or subcontracted to a consulting firm. Vendors selling the IaaS service should not be doing the pairwise comparisons.
References
1.D. Kocaoglu, “Hierarchical Decision Modeling”, Engineering and Technology Management Department, Portland State University, Portland, 1987.
2.A. K. Sadhu, “Delphi technique”, Managementversity, 29 November 2014. http://managementversity.com/delphi-technique/. Accessed: 10 March 2017.
3.P. Mell and T. Grance, “The NIST definition of cloud computing”, Technical report, National Institute of Standards and Technology, US Department of Commerce, September 2011. http://nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-145.pdf. Accessed: 16 March 2017.
Chapter 3
Technical Transformation: Cloud Computing
Amit Pingle* and Tugrul Daim*,†,‡
