Make stronger business decisions with advanced analytics

Part 2 of 2

Senior Managing Consultant, IBM

The first installment of this two-part series examined the core elements of focusing on business decision-making as a key pillar in the analytics conversation. In this concluding post, we’ll look at the other two pillars: embracing an agile culture and investing in an ecosystem of talent

Embracing an agile culture 

How should you continuously control variability and codify the decision logic of the best decision-makers, regardless of geography, domain or business unit? Can you continue to adjust the process in which improved decisions must take place? Is there a way to embed the iterative nature of solution development?

Advanced analytics initiatives have an inherent element of uncertainty. For instance, optimization models may not be able to converge on an ideal solution and predictive algorithms may not be able to achieve a desired level of accuracy. The right solution will differ by company and by project, and will depend on data and experts’ availability, the amount of the project's budget that can be spent on experimentation and on a company’s ability to absorb any cultural change resulting from the project.

To be successful, advanced analytics initiatives require a new project delivery approach that is akin to a scientific discovery, where an iterative, exploratory process is used to determine and implement the best possible analytics solution. For example, a recent IBM client described its goals as reduced power usage (which was the primary operating cost component) and fewer quality-related charges. Having clear and financially quantifiable objectives, the team made effective modeling design decisions while abandoning or trading off lower-level objectives that impeded the overall goal. The financial nature of targeted benefits have also allowed quick “go/no-go” decisions when assessing the benefits of model changes versus the costs of making them. It is also helpful to start with a simple model and continue enhancing the algorithms in iterative cycles to deliver increasingly valuable analytical capabilities, in a manner similar to agile development methodology. 

Investing in an ecosystem of talent

Is there an ecosystem of talent available to implement and embed analytics into the organization? When does it become necessary to train, hire or get external help?

IBM experience has shown that a project team organized around outcome-based activities rather than expertise clusters is more effective. Although this way of organizing the team can at first create friction between members accustomed to different work styles, productivity gained over the life of the project justifies the initial challenges.

Analytics talent management includes these three components:

  • Center of Excellence (CoE): A small group of resident experts with business domain knowledge and stakeholder relationships should be responsible for all analytics projects in the organization. In addition to providing continuity and consistency in project delivery, this executive team should support the business case, quantify benefits and leverage early adopters and cultural transformation.
  • Innovation and implementation partner: An external organization with industry best practices, research credentials and scale in technical capabilities should be engaged on a program or project basis to co-manage and staff advanced analytics implementation. They also should offer industry and domain-specific turnkey solutions to enable faster business value.
  • Data and technology experts: These resident technologists should understand the data sources intimately and know how they can be deployed to implement a project anywhere in the organization.

Advanced analytics solutions take advantage of complex mathematic algorithms to attain a level of analysis not possible or practical to achieve using conventional analytic methods (for example, simple regression or correlation analysis). They provide consistent and reliable decision-support capabilities that improve speed, quality and profitability in areas such as customer service, product design, investment decisions, equipment reliability, assets utilization, workforce optimization, infrastructure planning and safety improvement. Advanced analytics leads to enhanced decision-making that helps deliver a swift and high return on investment (ROI). Market leaders are embracing these opportunities as advanced analytics initiatives are implemented at a growing number of companies. Whether the analytics maturity of the organization is Spectators, The Pack, Joggers or Front Runners, by focusing on the three strategic pillars discussed in this series, the organization will be well positioned to enable better decision-making.

To start your advanced analytics journey today, visit this IBM advanced analytics resource page.