Let’s face it. With the volatility of the market and increasing uncertainties that arise within your business, you need actionable insights to contend with competitors buoyed by digital transformation efforts. The valuable data stored in your ERP system can help you make informed decisions, but that data is often the toughest to unlock. Getting the insights you need is a slow, complex, and expensive project – and your BI tool isn’t delivering because it is only as intelligent as the business acumen your organization can provide to it.
Over the last few decades, we’ve seen at least three distinct generations of BI technologies introduced to the market:
Products like Crystal Reports, Brio, ProClarity and Siebel Analytics primarily offered thick-client reporting solutions to replace the historically manual production of “paper reports” ubiquitous in every company.
Many of these products are amazingly still widely in use in today’s enterprise – multiple decades later – although we see a speedy migration to the third-generation BI solutions.
These early pioneers gave birth to the next generation of BI tools and, in many cases, were acquired by them over time. Solutions such as BOBJ, Cognos, and OBIE adapted to the requirements of the larger enterprise, introducing rich semantic models, governance capabilities and targeting a far larger audience inside the enterprise by providing capabilities for analysis, pre-built reporting, and automated refreshing, etc.
These solutions led to large-scale adoption in the enterprise, which led to most of the major vendors being acquired by enterprise software providers like SAP, IBM and Oracle.
Widely known as “self-service BI,” as it sought to break down the bottlenecks created by many Gen 2 solutions that relied heavily on IT services to deliver insights, solutions like Qlik, Tableau, Domo, and, more recently, PowerBI focus on providing the ability for end-users to create powerful and interactive visualizations, without needing IT to do a lot of the “semantic modeling work.”
Or at least that is what users of these solutions believed. The reality is the workload just moved down to the end-users.
With all the potential delivered by each successive generation of BI tools, why is it that so few organizations have begun to meaningfully tap into the most interesting, high-value, and even game-changing opportunities that BI promises – like decision intelligence (including AI, ML, and process mining), automated storytelling, or augmented insights?
It’s simple – the famous “80/20 data science dilemma” explains it best. Armand Ruiz explained it this way in a column for InfoWorld: “Most data scientists spend only 20 percent of their time on actual data analysis and 80 percent of their time finding, cleaning, and reorganizing huge amounts of data.”
It’s such a well-known problem that dozens of companies have started up aiming to automate even just one part of the process. Vendors like Matillion, Fivetran, Trifacta, Alteryx, and many more are taking worthy and exciting approaches to solving the problem using techniques like self-service data wrangling, ML and pattern recognition models to automatically discover semantics or converging visual data discovery with data prep to accelerate the users understanding of the data.
But these approaches will only ever yield modest improvements – particularly for the most critical data in your business – your ERP data. Your ERP is the system that allows your business to turn raw materials into products, products into revenue, candidates into employees, orders into invoices, and much, much more. The previously mentioned approaches are only helpful IF you can connect the insights from these new sources and techniques to your enterprise data, your production, your selling, your customers, your invoicing, your orders, etc.
BI can do a lot, but it is only as good as the ingredients you put into it. Where is the “business knowledge” that is essential to combine with the data and the systems to arrive at insights and knowledge? It sure isn’t happening in the visualization layer! What you need is both a view of what is happening in the cockpit of the plane you are flying AND a flight plan. In other words, you need real-time reporting and deep business insights to provide continuous intelligence for your enterprise. And you need that intelligence to feed your BI tool.
In order to do this, you need a business data model that abstracts the complexity of the underlying ERP data schema into user-friendly business terms. When used as a data source for BI, such a model presents complex ERP data in a more consumable and understandable fashion, freeing BI reporting development from the significant complexity of ERP data structures, speeding development projects, and making BI reports and dashboards more relevant and more intuitive.
Learn how Magnitude Angles applies the principle of a context-rich business data model to SAP and Oracle ERP systems, providing self-service operational reports from a library of more than 1,600 customizable templates, and BI integrations to provide true business intelligence for your business intelligence.
A version of this article was originally published by OATUG.org