Putting Analytics before Data to Build a Secure, Intelligent and Connected Enterprise

Share:

As the world emerges from the pandemic, one thing has become clear: the digital economy and its role in commerce will continue to expand. Organizations are acquiring digital assets at an accelerated rate, yet mining data for insights and securely integrating information across the enterprise remain challenges.

The success of an organization in the digital space will largely depend on its ability to combine technology and business strategies and lay the foundation for a secure, intelligent, connected enterprise (SICE) in a way that resolves this “digital gap.”

When thinking about analytics in the context of becoming a SICE, companies often focus on the need to balance connectivity (the integration of data from disparate systems) and security. This is because every new data element that needs to be exposed for analytics must be evaluated for potential exposure to cyber threats as well as for compliance risks. Enterprises set ambitious goals, including a vast and powerful data architecture upon which analytics can thrive. Theoretically, this is the right approach that will yield positive results over time. In the short term, however, balancing connectivity and security can put a drag on SICE ambitions and derail an overall worthwhile effort.

So what goes wrong?

Culturally, analytics lives in a different realm from IT and cybersecurity. Successful analytics teams are successful because they move at the speed of business and deliver quick results within an acceptable margin of error. The challenge here is, no margin of error is acceptable when it comes to security. As a result, organizations can end up in a situation where the two are hard to reconcile.

To illustrate, creating a fairly accurate customer valuation model with today’s powerful and ready-to-use machine learning tools is a straightforward process. The biggest stumbling block is most often data, which can be either lacking in integrity or unavailable at the level of detail or update frequency needed.  Barring these common data problems, a good model can be built in a matter of weeks. However, if proper data channels and security networks are not established, a project can get delayed for months. This is because, for most organizations, there are simply too many unknowns in building cloud-based analytics at larger scale.

So how do we address this issue?

Get specific with your data for the sake of analytics

If you ask an average analytics professional what data they require to complete a project, you may get a specific list of requirements with clear prioritization and scope. However, in reality, there is a strong inclination to ask for as much data as possible, just in case. This tendency to “boil the ocean” and capture as much data as possible is a good strategy in the long term, but, in the short term, it creates unnecessary complexity, most notably for security.

Analytics and security should work to identify data that fit the following characteristics:

  • The data should fulfill at least 80% of the goals of the analytics project
  • Data should not be sensitive or confidential
  • If you cannot avoid using sensitive data, focus on data that is available in a secure environment or won’t take a lot of work to establish in a secure environment.

If an analytics project requires a lot of data that cannot meet the above criteria, it is likely not the best candidate for your initial SICE ambitions, and you may be better off selecting something with a smaller security footprint.

Balancing smart and secure analytics

When it comes to security, there is little that can be compromised on, but analytics can adjust to drive significant benefit with partial results. Even the best intelligence is worthless unless you actually put it to use making intelligent decisions. Faced with this fact, it is important to be creative in how you accomplish smart and secure analytics. Here are some examples:

  • If gathering personal data on your customers in a cloud environment is a challenge because of security, limitations on legacy systems or regional compliance pressures, look at hiring a provider to manage some of this work. They may also bring additional experience to the table to help with the effort.
  • If your current cloud enviroment security architecture is too flimsy to support a more sensitive data intelligence project, consider spinning up a temporary architecture with security that perfectly fits your needs.

It is important to note that data quality and availability issues remain the biggest problem for most organizations when it comes to getting quality insights from their analytics. Ultimately, to support your SICE efforts at scale, broader security and connectivity transformations are necessary. The point is that those take time. And with boardrooms having short attention spans, it’s beneficial to show results early. Creating an analytics and cyber partnership early and prioritizing the right projects will get your teams focused on the right goals.

True business intelligence does not need infinite data to show results. By putting analytics first and not letting “better” get in the way of “good,” companies can prioritize their goals. This stepping-stone approach can help enterprise leaders set longer-term goals while driving short-term results that show the true potential of a secure, intelligent and connected enterprise.

ISG helps companies plan their SICE future. Contact us to find out how we can help.

About the author

In ISG, Olga supports our advisory capability in developing digital solutions with special focus on data and analytics. Olga’s extensive knowledge of analytics and data engineering framework combined with hands on experience in complex transformational projects results in unique insights invaluable for effectively assessing the data analytics solutions for ISG’s clients.

Share: