Analytics Knowledge Management – Why does it matter?
Business decision-making, in today’s world, benefits from availability of massive amounts of data and a wide range of analytical tools to derive insights from it. Tools and automated processes for capturing, storing and preparing data provide access to torrents of data that organizations can leverage to drive business decisions.
However, the wealth of data and the speed at which it is growing has only exposed daunting challenges in gleaning meaningful insights for decision-making fast enough to be actionable. Machines can do the grunt work of data plumbing and to some extent generating automated insights with the advent of AI, but the human element still plays a big and indispensable role in truly generating value out of data.
The Typical Process of Running an Analysis
Extracting insights from data is an iterative process of discovering and exploring data from multiple sources, blending and transforming it and analyzing it to find relationships and patterns that can be made actionable. Oftentimes, this process also involves collaborating with other analysts, data scientists, data engineers and subject matter experts to get access to the right data, understand it, trust it and leverage it in the analysis.
During this iterative process of data discovery, data preparation, collaboration and analysis, raw data goes through several transformations and is augmented with knowledge and expertise from across the organization to analyze it and interpret the results to make actionable recommendations.
In most organizations today, all the knowledge that goes into an analysis typically never gets formally recorded in a form that is easily searchable and reusable in the future. Such knowledge includes any of the following and more,
- Data transformations including filtering of outliers, sampling, calculations, blends, etc.
- Data interpretation, which generally involves working with the data owner or subject matter expert
- Queries, statistical models and code
- Results, recommendations, reports and dashboards
All this knowledge, along with the data itself, constitute the Analytics Assets that an organization possesses. However, in the absence of a way to capture these in a searchable and reusable format, the assets are prone to getting lost or buried in places like an analyst’s laptop or simply in their memory. What happens then if an analyst leaves the organization? All the knowledge that was created is essentially lost. The inability to reuse past knowledge and analytics assets invariably leads to duplication of work and reinventing the wheel.
Benefits of Analytics Knowledge Management
Organizations build analytics knowledge over time. Knowledge is created with every analysis, every new data source and with every business question that gets asked and answered by leveraging data. Here are some of the benefits of robust analytics knowledge management.
1. Reuse of assets for shorter time to insights
Consider an analysis that combines two siloed datasets, one containing transaction data and another containing customer data associated with those transactions, to gain insights on purchase patterns by demographics, geography or some other customer attribute. Creating this blend of two datasets involves discovering the two datasets, understanding the data that’s in them and what column to blend them on. Now consider another business question that comes in at a later point that seeks to understand purchase patterns by various customer attributes during periods or certain weather events, which of course is a third dataset. If the analyst trying to answer this question could search and find the dataset that was previously created using transaction and customer data, their analysis could simply reuse this dataset and only blend in the new weather data with it.
Without knowledge management, or, in effect, without access to this previously done analysis, the analyst would have to discover all three datasets from scratch, understand them and blend them to use in their analysis. Having to redo work previously done naturally increases the time required to answer a business question, which can have a tangibly negative impact on the agility to make data-driven business decisions and consequently on business performance.
2. Completeness of Analysis – solving for the Unknown Unknowns
Consider the same example as above, where an analyst blends in weather data with transaction and customer data. It is not uncommon to have 3rd party data, such as weather data, in the form of an extract from an external vendor. The same holds true for data that may come from a source that is not directly connected to the analyst’s BI or Data Science tool of choice or available in an easily accessible data warehouse, thus forcing them to use an extract of it from another system or SaaS application.
If this data and the knowledge about it is not captured in a centralized, searchable format, it is quite likely that another analyst trying to answer a business question in the future that could have benefited from the use of this data may not even know that such data was available, or even worse, think about leveraging such data to answer the question at hand. This could lead to an incomplete, possibly misleading, analysis and recommendation.
3. Fostering Collaboration for knowledge and expertise sharing
Collaboration between not only analysts within the same team, but also those across business functions can lead to better analytics driven by the sharing of knowledge and subject matter expertise. A question that analysts frequently have is who to reach out to if they had a question about a data source, a column within a table or how a certain analytics asset could be used in an analysis. In a world without analytics knowledge management, the process involves emailing a bunch of people or walking around the office asking co-workers who might know the answer. With a comprehensive analytics knowledge management solution, the analyst could simply search for available data or analytics assets and the system would not only list them out, but also provide information about who owns the data or asset, what projects has it been used in most recently or most frequently and who the analysts are that own and use it. The analyst could then reach out to the right person within minutes of the search and start collaborating with them. Leveraging the knowledge and subject matter expertise of colleagues that own or have worked with an analytics asset before can not only significantly accelerate the time to insights, but also provide invaluable inputs for a better, more comprehensive analysis.
How does Neebo enable Analytics Knowledge Management
Neebo is a SaaS solution that, among other things, offers comprehensive, AI-augmented analytics knowledge management capabilities that let analytics teams build knowledge over time and access it seamlessly for use in future analyses, without ever requiring to duplicate efforts or trying to figure out where, or with whom, the knowledge resides. Neebo also enables collaboration between analysts by letting them work together on an analysis in, what we call, a workspace, which is a collection of all analytics assets required to answer the business question at hand. Learn more about Neebo’s knowledge management and collaboration capabilities by scheduling a 1:1 Demo.
In addition to Analytics Knowledge Management and Collaboration, Neebo also lets you connect to any data source virtually, be it on-premises or in the cloud, blend and transform data to create new data assets without having to worry about the physical location of the underlying data sources and consume them in any Business Intelligence or Data Science tool of your choice. Click to Learn More.