For mathguy18 ONLY - Business Analytics Implementation Plan Part 2

Gradient must have a clear understanding of some basic factors so as to be able to ensure a successful implementation of the BA. These factors include the organization structure, mapping business objectives, the type of data required, and then the implementation itself. For Gradient to ensure successful implementation it must assess the right number of people required for implementation (Bartlett, Randy, 2013).

2.1 Organization Structure

It is required that the core team comprises three people: an analyst, the developer, and a statistical modeler. A fully functional team should also include a project leader, evaluators/testers, and business analysts. One challenge that Gradient is likely to face is the modality of finding and budgeting for the resources. Sometimes it is hard to locate competent experts such as statisticians and developers who are specialized in business analytics in the marketplace. The success of the BA implementation will heavily depend on the level of skill of the persons engaged in the implementation and so Gradient must ensure that it locates all the best of these.

In the case where the company has industry-specific knowledge and experience on statistical modeling then outsourcing the BA is an efficient and cost-effective solution. There are quite a number of approaches that Gradient can use so as to be successful in using this approach (Stubbs, Argosy 2015).

1. Start from a Proof of Concept (POC) and, from here, engage larger projects. Under this approach the BAs are implemented phase-by-phase so that if the pioneer phases are successful then we can engage bigger projects. This approach has the merit that it allows for time to study the project phased implementations and take corrective steps so that the large errors are not carried forward to the large projects.

1. Ramping up with the outsourced company followed by a move to bring the BA area into the company’s full control. This approach also will enable Gradient to study the characteristics of the outsourced company prior to bringing the analytics to the full control of the outsourced company.

1. Full outsourcing for the long run and then using a model for revenue-sharing. This approach has the risk that if the outsourced company is not a competent one in the field on which it is outsourced then the business can run the loss as a result of the company’s incompetence.

For Gradient which has an elaborate company structure, it could be preferable to use staff augmented model for the analytics. It can form another challenging budgeting for these resources and the BA software. This results for supply that exceeds demand in the market thus pushing the costs up with the effect that the company pays a premium for them. So as to justify these resources there is the need to show a return on investment but this can be challenging in an area where there is adequate and or elaborate historical experience. However, the best way of approaching this would be to start from a well-defined mini-project and then do an extrapolation from the results of the whole project’s ROI. The other potential solution to this issue is outsourcing (Saxena & Srinivasan, 2013).

After having a proper understanding of the organization structure then Gradient to specifically map the core objectives of the business based on the questions that it would want to have addressed. This requirement is usually overlooked due to its perceived simplicity because of the way in which it recurs. Gradient should focus on knowing how they can use predictive analytics in lieu of defining the objectives for which they are seeking solutions.

Having a clear business understanding of the issue at hand is a fundamental step in implementing the business analytics. BA could be based on a blend of mathematics and business knowledge which should have the necessary precision to bring a solution to the specific question (Davenport & Harris, 2007).

It is upon the leader of the business analytics team to see into it that the best practices are complied with starting from mapping out the business objectives to the question that needs to be addressed and which the analytics are supposed to provide to. The business analytics leader has the responsibility of identifying the success metrics that are used in the evaluation of the project. This measure should be in business terms for example the potential revenue or costs saving that the project has for the entity (Davenport Thomas , 2006).

Data availability and quality are the two key issues to which the results of the business analytics have direct bearing. One way of dealing with data quality issues is to ensure that they are dealt with the issues prior to undertaking the analytics project. It is very difficult or even impossible to ensure that data used for the analytics is perfect (Bartlett, Randy, 2013). However, the project leaders must make an agreement upon some allowable limits that the project outcome should be. Besides, the project team can use a number of statistical techniques for refining the results and to minimize issues that may corrupt the quality of data. One such technique could be to segment the results into “expected results” and “unexpected results” for the purposes of evaluation (Bartlett, Randy, 2013).

The issue of data availability could sometimes prove more complex due to issues related to extrapolation results especially when data is not available and further when it is not possible to procure additional data (Davenport Thomas , 2006). For example, Gradient could make an extrapolation of the results for an ethno-demographic business promotion projects into stores because of data limitations but that would still fall into the same classification as though there were sufficient data.

Validating the extrapolation will be based on classification or segmentation is valid or otherwise. For example, a mini store (in terms of sales volume) in Miami may not exactly be comparable with another mini store in Southern Carolina because even though both of them cater for Latino customers, the population characteristics may differ. On the other hand, to extrapolate the results within a given geographical area of Miami, a segmentation of a mini store could be useful within the geographical area in mini stores without data available.

Understanding data is an essential step in the implementation of business analytics because it is the only way of, and juncture, at which the project specialists ensure that the organization data and the business analytics can communicate. At this juncture, the compatibility of the two items is enhanced. Either the new system has to conform to the organizational data or vice versa. Whichever way, compatibility must be ascertained to ensure that there are no disagreements between the two. The direction of compatibility would be determined by the availability of data or the manner in which they are arranged, or are required to be arranged.

Implementation is a vital process since it is the step where transformation of data is done to generate meaningful information. In order to maximize the return on investment, it is essential that one has due understanding that business analytics have varied meanings within the same company (Saxena & Srinivasan, 2013). The executives of the firm may be interested in knowing how the tactics used in developing the business analytics are aligned with the company’s strategic goals. On the other hand, line managers would be interested in knowing how to accomplish the specific line management goals such as monthly objectives and weekly objectives, and so on. The employee who is on the field has short targets and he would want to know the way in which he can accomplish today’s target (Saxena & Srinivasan, 2013).

When the results of the business analytics are visualized through the dashboards, the decision makers are able to catch a quick and firm grasp of the meaning of the information available to them. Gradient should establish the way it would want the information layer to be presented to it prior to embarking on the business analytics project. It is required that the information layer be correlated to the objectives of the business (Davenport Thomas , 2006). Further, there is need for the information to be flexible so as to allow for the addition of new requirements. Gradient should put into consideration utilizing the advances in visualization performances at the time of planning a dashboard.

As Gradient Connections attempts to embrace and also streamline business analytics projects into its operations, it should plan to confront issues that can overturn its goals. In the past few years it was commonplace to hear of 50 percent IT project failures. One of the most important lessons to learn from that is that limiting project scope, resource budgeting and project-planning or of great essence in business analytics planning. When planning analyzing projects, it is important that companies incorporates a modified version of the lessons learnt from such failures. Company and project leaders need to think big and at the same time start with measurable projects. Also, they should avail themselves of best practices and identify issues that can derail the project in early and to use the experiential lessons to ensure that the project succeeds.

References Bartlett, Randy. (2013). A Practitioner’s Guide To Business Analytics: Using Data Analysis Tools to Improve Your Organization’s Decision Making and Strategy. McGraw-Hill. ISBN 978-0071807593. Davenport Thomas . (2006). Competing on Analytics. Harvard Business Review. Davenport, T. H., & Harris, J. G. (2007). Competing on Analytics: The New Science of Winning. Harvard Business School Press. Saxena, R., & Srinivasan, A. (2013). Business Analytics: A Practitioner's Guide (International Series in Operations Research & Management Science). Springer. ISBN 978-1461460794. Stubbs, E. Argosy University, (2015) Delivering Business Analytics: Practical Guidelines for Best Practices. Retrieved October 18, 2015 from http://myeclassonline.com/

10