Building the right data analytics solution for your business can really pay off, but it requires consideration and planning to build a strong, yet flexible platform.

You want data analytics that not only answer your first set of questions, but also the questions you haven’t thought of yet. You need business analytics to account for the future, without knowing what all the factors are. You can create a sustainable analytics platform – one that is adaptable and extendable. This post will go through the cornerstones of building an accurate, vibrant data analytics solution, based on CMA’s 20+ years of experience in data management.

Cornerstones of Creating an Analytics Platform to Answer Future Questions

The key to planning is to not plan everything at once. It may sound counter-intuitive, but segmenting your planning allows you to go through the process in an incremental way. Getting your cornerstones in place will begin to yield surprising business intelligence, and once you are ready to scale up, you will have a better handle on which questions to ask.

The following cornerstones don’t occur in any particular order, but need to be addressed simultaneously for best results.

• Consider your first round of questions
• Tackle and trust your data
• Evaluate data platforms
• Recognize a comprehensive information management framework

Consider Your First Round of Questions

There’s nowhere to start but at the beginning. Odds are, you are delving into data analytics for a few specific reasons, so start there. But you need to realize that this first round of questions, phase 1, will probably only cover about 20 percent of the overall questions that need to be asked. And that’s okay.

But to make the process work, your organization must also have executives predisposed to fact-based decision making. A “data-allergic” management team that prides itself on making gut-based decisions is unlikely to be supportive. Any analytical initiatives in such organizations will be tactical and limited in impact.

But reviewing what analytics are being used in “sister” agencies can be a great starting point, but don’t let it be the end. Every workplace is unique. Develop a conceptual set of use cases or analytics within your agency. This list of questions will evolve into phase 1 of your analytics solution.

The next step would be to figure out what data you need to answer them.

Tackle and Trust Your Data

When a business considers initiating analytics, trust is paramount. You need to have consistent, quality data for decision making. This can first require improving your transaction data environment. If a company or institution has poor quality data, it should postpone plans for use of analytical processes and fix its data first.

Elevate the quality of your data first. Organize your master data. Form data sharing agreements with peer organizations and consult freely available sources of data. In recent years, data management (also known as information management) has evolved to a discipline that manages data in a holistic fashion, and includes:

Operational/transactional systems to support the day-to-day needs of the business
Decision support systems to support requirements for trend and historical data analysis
Data movement and integration mechanisms to support real-time, near real time, and batch data transmission and consumption requirements
Information delivery methods to support reporting, data feeds, and related data consumption needs
Many times, organizations will rely on their chosen analytics tool to set the parameters of their data. You need to go beyond the analytics tool. If you just ask the tool to automatically set everything up, you won’t get the results you need. There is no magic. You need to examine everything and set up your scrubbed data for scalability.

Evaluate Data Platforms

Analytics tools can be costly, and are even more costly if they go underutilized. Due diligence in product identification, assessment, and fit against your primary goals/outcomes is essential. Sometimes, management is focused on what their peers are using, or what they see on the internet or in the press. Sometimes, they will have prematurely invested in funding a tool that may not (and probably won’t) work for them.

There are many tools out there, which can be broadly divided into two categories: Business Intelligence (BI) and visual analytics. BI platforms include:

• Oracle Business Intelligence
• SAP Business Objects
• Cognos
• MicroStrategy
• SAS
• Hyperion
• Actuate

A new generation of analytical tools enable the manipulation of data analyses through an intuitive visual interface, such as geospatial. Visual analytics tools include:

• Tableau
• Salient
• ADVIZOR
• Qlikview
• TIBCO Spotfire
• JMP
• Visual Analytics

All these tools have strengths and weaknesses, which depend on your situation and needs. CMA has worked with all of them, to one degree or another, and we can help you sort through them.

But first, know that technological integration is often the least of your worries when it comes to setting up a new data analytics platform. And you can minimize these worries with a comprehensive process to build your analytics solution.

Recognize a Comprehensive Information Management Framework

How completely are you managing your data? Data is complicated and spread out. Data goes through various areas (technology, people, and process) and encompasses many domains, collaboratively and at the same time.

When building your data analytics solution, you need to have all these areas and domains considered and covered. Bring the relevant people and areas in at the beginning, and maintain strong partnerships all along the way.

These layers and components are all critically important factors. They need to come together to create a sustainable, accurate, and useful BI solution. Building a framework that considers all data factors will ensure your system fits in the future, and not just as a one-off situation solver.

Conclusion

Include these four cornerstones in your planning process to create a data analytics platform that is the perfect fit for your organization. Trust your data, ensure that your data management is comprehensive, know what questions you want answered, and evaluate your data platform accordingly, You’ll be well on your way to creating an analytics system that is effective and efficient.