From siloed analytics to collective intelligence

The evolution of Analytics and Business Intelligence

The last 15 years analytics and business intelligence have always been on top of executives’ agenda. Offering a powerful instrument that provides important and valuable information to support companies in their decision-taking, aligning finances with their strategic goals. Implementing a Business intelligence solution means cost savings, sales optimisation, increasing revenue and profit as well as insights in the best time for a new investment.

Industry has responded to this demand with an impressive growth of Business Intelligence providers (tools and services) in the market. Although this has a positive impact on the pricing, choosing the right tool, partner and approach became an industry on itself. This complexity has for sure its contribution in the fact that more than 60% of the analytical projects seems fail (according to a study of Gartner in 2019).

We do see the market reacting to this: 2019 was a year of transition toward cloud ecosystem dominance. The rapid growth of the Microsoft Azure-based Power BI cloud service, along with Salesforce’s acquisition of Tableau and Google’s purchase of Looker, signaled a change whereby cloud stacks are now expected to come with an even more competitively priced BI platform and an increased portfolio of functionalities and most important connectivity possibilities.

Siloed Analytics

"Excel sheets do an excellent job storing and arranging information for data analysis, but they will only tell you what you already know."

The complexity of implementation paths and tool selections has also led to a diversity of isolated implementations within organisations having teams taking own initiatives in setting up an analytical environment to support their own needs. From full-blown business intelligence architectures with ETL frameworks (Extract, Transform and Load), data warehouses with advanced cubes or universes up to the less sustainable, but often very powerful and Excel dashboards. And this makes perfect sense. We all see the added value a good analytical tool can bring and how it can support us in achieving our goals backed by relevant data.

Organisation-wide implementations of such systems are often projects that come with a certain cost and can run over more than one year. Holding risks as requirements will change. Especially in analytical and business intelligence projects. The implementation track needs to be set up from the beginning to be able to handle these changing requirements much more than in any other software development track as well as it needs to collaborate with existing initiatives on analytics, something that is very often overlooked. Designing a solution that fits all has also been proven to be a very difficult story.

Enterprise-wide business intelligence implementations will start creating a common set of measures, building a thesaurus and aligning semantics across involved departments. An exercise of building common definitions that fits all, or at least as much as possible. Of course, we want to avoid having multiple calculations for the same measure and in sake of maintainability we don’t want to have it stored in multiple locations either. Again, what we experience in this approach is that change, and the existing culture is often not being encapsulated or respected.

Maintaining decisions and calculations where they serve the organisation best is not a bad approach, even more if you don’t, local initiatives will flourish even more so why not supporting this from the start, more in a controlled way fitting into a bigger business intelligence picture? It enables much more rapid implementations and shorter change cycles, a vote for maintainability and we don’t need to serve a lot of different kings. Focus on business results rather than adoption rate in the beginning has been proven a good practice.

How do we make sure shared information and measures are exactly what others expect it to be and does this implies even more tools entering the organisation? Talking about the IT toolset maintainability and related license costs, can we still fit this approach in a strategy that strives for more consolidating and centralisation?

Collective Intelligence

“Shared or group intelligence that emerges from the collaboration, collective efforts, and competition of many individuals and appears in consensus decision making”. 

No matter how small your initial analytical implementation is, in the long run it needs to connect to enterprise applications and actually accelerate the adoption through the organisation. Without the added data from enterprise applications, business intelligence adoption tends to stall, stop and eventually decline. It’s a fair assumption to say that if any company wants to gain high business intelligence adoption levels, it’ll need to integrate with CRM, ERP and other enterprise systems that are core to their daily functions as a business.

Knowing which analytical initiatives already exist within the organisation and how mature these are, will deliver valuable information from within the organisation. Some local projects most likely already did a search to find the best tool for their purpose. And not only on the tooling, tapping into the rationale of those initiatives might broaden your vision and potentially safeguard you from pitfalls they already encountered.

It’s a thin line to walk in order to support these local initiatives and yet not to encourage people in moving away from overall strategy and analytical growth. However, don’t underestimate the value of what’s already there and the reasoning that went into it. More than 90% of the teams have some kind of analytics running, it doesn’t need to be an actual tool installed, it can be very basic excel sheet.

Making the tool selection using collective intelligence has been proven to have a positive effect on later adoption rate. Pushing a tool through an organisation often has the opposite effect causing more local and unconnected solutions pop up and live their own isolated life.

Think outside the box and dare designing an architecture that contains different setups serving the local analytical needs with local responsibilities as well as serving the overall companies’ goal can actually even save you money and holds less resistance.

What’s next?

While the analytical and business intelligence market is still in development new technologies like artificial intelligence and applied analytics are entering the field. Having your business intelligence data landscape in place with relevant validation rules delivers the foundation AI and applied analytics need as fertile ground to grow and flourish.

Start your BI integration strategy by concentrating on legacy and 3rd-party databases first before moving on to larger, more complex integrations. These integrations involve enterprise applications like customer relationship management (CRM) and enterprise resource planning (ERP). Integrating with enterprise-wide systems including CRM and ERP is where the value of BI increases exponentially.

Select a flexible, modular system that can scale with your user’s needs while respecting existing initiatives and lose the idea of replacing them all. BI adoption increases when a system can flex and respond to the needs of a broad base of users without forcing them to change how they work. The more modular and agile a BI system is, including the flexibility for defining custom workflows by business analysts using different tools, the greater the level of adoption will be.

Business Intelligence is a smart tool that will help you reveal tendencies in your past performance that could otherwise go unnoticed. You can identify significant trends in your data with the potential to unlock new growth opportunities. By analysing your past performance in context and trying to understand the factors that influenced the best or worst results, you can discover the key to future growth.The body content of your post goes here. To edit this text, click on it and delete this default text and start typing your own or paste your own from a different source.