Wednesday 28 January 2015
Reading Time: 3 minutes
Innovation can always come with a price, and being positioned so differently to other business analytics solutions search-based business analytics solutions can fail to pass the initial ‘looks like what we know’ test. Search-based business analytics solutions are revolutionary in the way they handle data sources with the objective of sub-second response times with enhanced data discovery techniques and reporting options. With Big Data only getting bigger the speed and efficiency of reporting is becoming a key factor in analytics procurement.
Having read a BI analyst’s assessment (someone who had only ever known EXCEL and old style reporting) of what they thought a search engine based business analytics solution would be like, I thought it was worth clarifying a few things…
1. Search can’t do OLAP style reporting and we need structures to be pre-defined
Search-based software by design operates on the raw data, rather than aggregated data structures such as an OLAP cube. OLAP cubes merely pre-filter data into set structures, whilst search-based solutions provide greater flexibility over index structures by being changeable in real time and not requiring reindexing of a data source when changes are made. Search-based solutions are compatible with OLAP data structures by deriving its own structure from the raw data and indexing process.
2. We need to have pre-canned hierarchies and dimensions so users can navigate where to go
Healthcare organisations are often spread over multiple CCG’s and it has been commented that search-based BI does not allow sets to be created akin to OLAP structures. This is incorrect – saved queries provide the same function as sets and are changeable in real-time by the end-user, rather than pre-defined in the back end. Multiple queries can be layered to create complex data structures and do not have to be re-entered every time a report is made.
3. Search engine data is always going to be out of date
Search-based solutions are iterative in nature. Reports do not update in real-time and have to be run to allow flexibility in end-user hardware whilst keeping sub-second response times across any size data set. Solutions that update reports in real-time encounter performance issues with only small to medium style data sets and have difficulties processing data sets surpassing a million records. The key advantage to search-based reporting is infinite scalability combined with sub-second response and query times.
4. You can’t answer complex questions without SQL
Modern search-based solutions use a point and click interface coupled with Boolean operators to make the ability to do SQL level queries available to anyone who knows how to use Google by merely clicking on a few records. From fuzzy queries to advanced calculations any desired query can be made additively by simply clicking on records in the front-end of the solution. Want A&E admissions between 3pm and 7pm on Fridays in the last week at North Staffordshire hospital? Either click on the time, date, hospital and department in the query screen (four clicks from your entire data set to actionable information) or create a Venn diagram that both creates the queries and delivers a visual representation of the information without the need to create any reports.
5. Searching gives you lists not dashboards and reports
This one is simple. Modern search-based business analytics solutions have all the reporting capabilities of other business analytics solutions whilst providing a service that scales across unstructured and structured data sets with unlimited records. It is not limited – moreover, it aspires to be without limits!
The key issue is with innovative takes on solutions being a risk and dissimilar to what users are familiar with. Search-based BI has been built from the ground up to be more usable and easy to learn than other solutions whilst providing functionality superior in every aspect from reporting to indexing, publishing to data organisation. Just because it’s raw data doesn’t mean it’s inferior to legacy OLAP structures! Give one a try and you’ll see for yourself.
CXAIR – Search-Based Business Analytics
CXAIR is a search-based business analytics solution designed to quickly and easily turn data contained in disparate systems across the organisation into actionable information from within a single application. By using CXAIR, users at any level of an organisation will gain immediate insight within their function or business area and be able to make informed decisions and take action to improve both the quality and performance of the services being provided.
In a dynamic health environment with many challenges and new requirements, CXAIR has been proven to help the NHS deliver improved services quickly, easily and at a reduced cost compared to other analytics solutions. In addition, ad hoc analysis requirements can be supported and quickly delivered alongside standardised reporting without requiring additional resource or diverting existing resource from value added analytics to support these requests.
CXAIR is designed to ensure that everyone is able to securely access information relevant to their role; delivered in a self-service format that works for them – including many advanced visualisation capabilities such as interactive Venn diagrams and geographical analytics using the familiar Google Maps interface.