A Data Science Central Community
Data discovery’s next chapter will see an explosion in analytics globally – from tens of thousands of data analysts today to tens of millions of business users within five years. Key drivers will be further advances in easy-to-use interfaces, the marriage of data discovery and unified information access (UIA) and greater mobile capabilities. This is in keeping with the shift in enterprise spending from legacy database architectures with rigid historical reporting to real-time, event-driven, in-memory analytics platforms that speed and improve decision making.
As a result of their flexibility, these tools are also now taking a lead in providing a direct integration with big data platforms, such as Hadoop and Cassandra. They are distinguished by beautiful presentation interfaces built on top of a new infrastructure foundation that help non-technical users see and interpret data more easily. Established vendors are addressing this gap in the stack with new products coming to market. While the large vendors continue to dominate overall BI platform market share, the momentum experienced by data discovery vendors marks the beginning in a long-term shift in market share that will also destroy the pricing umbrella enjoyed by the legacy BI solutions.
A principal value proposition of data discovery tools is that users can deploy the technology without dependence on IT in a matter of days or weeks, as opposed to up to 18 months for traditional BI tools. This faster path to user productivity – with current, real-time data as opposed to stale, inflexible data sets – is a key differentiator in lowering TCO.
The dawn of self-service analytics
As opposed to traditional BI solutions which are built on hierarchical query of pre-aggregated data sets defined by IT, data discovery tools put the entire data set into RAM to allow users to interact with data and make inquiries based on the way their minds work. The result is more insightful analysis that is unique to each user’s specific needs.
As the cost of RAM declines with the pervasiveness of 64-bit computing, in-memory analytics will become feasible for many organizations. With newer in-memory OLAP architectures, users can run far more sophisticated data analytics applications in real time that they could with traditional multidimensional OLAP architectures using conventional relational databases.
Facilitating access to a wide variety of data types without restrictive metadata layers against which users can conduct real-time searches and drive deeper, more valuable insights will make the ROI compelling. Users can search associatively and define and create visualizations of the data in formats they prefer. This user-defined approach to analytics, breaks down one of the barriers to adoption of traditional BI solutions and open the market to a much larger potential user community.
As these users seeking on-the-fly data feeds and statistical analysis, expect newer capabilities of BI platforms to focus on predictive analytic models and forecasting algorithms that can be more easily consumed in dashboards. Also expect to see the integration of UIA, which combines the best of both search and BI capabilities. UIA software allows users to create a mini instant data warehouse by uniting access to multiple types and sources of both structured and unstructured information in a single repository called a database table. Users can then conduct searches and generate BI-type reports.
UIA technology will replace large legacy enterprise search apps because of its ability to provide a single view across all information. Another related shift will be the repurposing of traditional data warehouses to eventually be used only for data that is not queried frequently.
Finally, mobile BI has the potential to significantly expand the population of users. Over the next year, we will see significant advances for facilitating the delivery of BI applications to a more user-friendly and mobile device.
For vendors, a three-pronged go-to-market strategy
Business users are becoming increasingly influential in the purchase of data discovery solutions, with easy-of-use surpassing functionality as the primary BI purchase criteria. Many are abandoning notoriously complex and inflexible traditional BI platforms and bypassing IT in the purchase of data discovery tools that offer a faster, easier and more efficient ways to model, navigate and visualize data in their vast data stores.
While this may risk the creation of fragmented data silos, it has also driven a dramatic expansion in the average number of users per deployment. IT ultimately gets involved by providing architectures, methodologies and information governance policies that bridge the gap between legacy BI and data discovery solutions.
The go-to-market strategy for data discovery vendors should consist of a three-pronged approach that addresses specific use cases or pain points (read: solution selling). Vendors should be flexible in how the customer wants to consume the product, whether it be in the cloud or on-premise. With this in mind, the following strategies extend from large enterprise-wide solutions to departmental deployments and small and mid-sized businesses.
This balanced, flexible approach will drive faster adoption by broadening market coverage. It will also help buffer the operating margin impact of ramping distribution and customer acquisition.