If you were to climb Mt. Everest, you would face many dangers, including large crevasses in the glacier. Without best practices and a phased ascension there is a large probability that you’ll get into serious trouble and fail.

When it comes to updating your data and analytics systems, the challenges can look equally difficult. To cope with changing and growing business needs, the obsolescence of incumbent technologies, and the pressure to reduce costs, IT has no choice but to evolve and change its underlying architecture for analytics. The process could be like a journey to the top of the world, or it could bring many pitfalls.

There are basically two strategies for updating your data and analytics infrastructure: augment your incumbent technology or replace it with new disruptive technologies.

On the surface, replacement looks appealing bringing the promise of a new, modern, clean and up-to-date architecture. But it has many risks, including a long, tunneled project approach, less mature technologies, and a lot of change management for the business.

One customer I work with embarked on a replacement project with the goal of exchanging the company’s data and analytics architecture with exclusively open source technologies. After several years, the project failed, partly because the business users rejected the long delivery delays and the changes brought about by the new solution.

On the other hand, augmenting existing technologies allows you to reuse what is working well and worth a lot. Augmentation can also offer a phased approach with clear goals, change management best practices and less risk through a mix of proven and new technologies. I defined this process as modernization in a previous blog post, Three ways to improve your analytics architecture. Experienced technicians in the field tend to recommend modernization to minimize risks and cost.

A continuous process of data and analytics modernization following the life cycle of your business problems can help you address new and changing business goals. To kick off a modernization project for data and analytics, and to determine your best path forward, I recommend building these three deliverables:

• A target architecture: Assess the challenges of your current data and analytics architecture, and design the target architecture using three ways to modernize and expand your analytics programs.
• A roadmap: Design a phased approach for modernization with short periods, quick wins and clear criteria of success for each steps.
• A business case: Define the total cost of ownership for the existing and targeted data and analytics environment on a given period of time, and clearly outline the savings between the two.

In Mt. Everest, some of the crevasses are so wide that three or four ladders are tied together to make the crossing. Because the ladders move with the glacier, highly experienced Sherpas work every night to maintain the ladders, tightening them or moving them to a better position.

Likewise, the three deliverables described above will help you determine the best journey for evolving your data and analytics architecture, which can be maintained continuously by your favorite data and analytics partners, like SAS and its ecosystem. If you want to know more how SAS can become your trusted technology and business partner, please look at SAS Consulting service offerings.