January 12, 2018
Most business intelligence (BI) applications are constructed using point-and-click drag-and-drop graphical user interfaces (GUI), but still require end user training and long ramp up times, although not to the extent of learning and writing SQL code. Few business decision-makers access data directly, instead rely on trained business analysts, even for basic information. Modern BI tools have sophisticated and often complex instrumentation (for drilling, pivoting, sorting, ranking the results), and require users to be at least somewhat familiar with the data sources and business metadata. This poses a barrier for wider adoption of BI for different levels of users.
Is there a better GUI for BI? You betcha!
Executives and non-data analysts have always wished for data to simply “tell them a story”. This has largely come true with natural language processing (NLP) and natural language generation (NLG) technologies (we’ll leave NLG vendor landscape for a different discussion). Users can now ask a question in a natural language (where NLP translates a question to a query, aka text-to-query) and get an answer via a programmatically generated narrative based on the result set returned by the query. The NLG narratives are especially effective when displayed side by side with a visualization. In addition to NLP and NLG capabilities built into BI tools, some BI providers are also creating chatbots as separate applications. These can allow non-technical BI users to ask questions and receive dynamically generated data visualizations and written highlights without knowing anything about the underlying data structures or metadata.
Here’s the current landscape of BI platforms with search like GUI
- BI generalists with search like GUI as an option. Please take a look at the our latest evaluations of leading BI platforms https://www.forrester.com/report/The+Forrester+Wave+Enterprise+BI+Platforms+With+Majority+OnPremises+Deployments+Q3+2017/-/E-RES129453 and https://www.forrester.com/report/The+Forrester+Wave+Enterprise+BI+Platforms+With+Majority+Cloud+Deployments+Q3+2017/-/E-RES137263 . Specifically, take a look at Figure 2 in both of these documents, the Current Offering “NLP and NLG” line item. A score of 4 or 5 is a good indication that a vendor has a solid search like GUI capability. We called out BOARD International, GoodData, IBM Watson Analytics, Information Builders, Microsoft PowerBI, MicroStrategy, Oracle, Qlik, Salesforce, SAP, SiSense and TIBCO Spotfire for having robust offerings in this area. Tableau has recently acquired ClearGraph which also significantly boosts Tableau search like GUI functionality.
- BI specialists with only search-like GUI: ThoughtSpot and AnswerRocket
- Search tools with some BI capabilities. From this list https://www.forrester.com/report/The+Forrester+Wave+Cognitive+Search+And+Knowledge+Discovery+Solutions+Q2+2017/-/E-RES136544 we often see Attivo and Squirro in the BI context
- Search/BI tools that seamlessly integrate structured/unstructured data: Attivio, Stratifyd, IBM Watson Explorer
- BI / text mining tools which go beyond semantically enriching data for semantic/cognitive search and analytics, but also attempt to uncover structured data relationships like primary/foreign keys, etc: Cambridge Semantics, ClearStory Data.
- BI chatbots: Oracle (Day-By-Day), Microsoft (Cortana), Looker, SiSense. Many other leading BI vendors plan to introduce BI chatbots in 2018