Will SAP HANA really ‘herald a new generation of analytics, business applications and IT simplification with SAP in-memory computing technology?’ Daneel de Villiers, managing director of the SAP consultancy, De Villiers Walton, evaluates the new technology from SAP and discusses the options available to firms striving for better Business Intelligence.

After some time in development, SAP now believes it has found an ideal solution for generating fast access to crucial Business Intelligence (BI). ‘HANA’ or ‘High-Performance Analytic Appliance’ is a new innovation from SAP that combines in-memory processing, optimised database structures and some innovative indexing, which together allow access to business analytics at unprecedented speeds.

In its simplest terms, HANA connects to existing SAP and third-party systems and creates real-time copies of the data. Once in the appliance, the data is compressed, transformed and indexed and then made available for analysis by business users, typically via an SAP Business Objects interface.

Access to essential business information is therefore extremely quick and is based on the most up-to-date information available. In theory, the functionality that HANA provides will make it far easier for the end user to access the information they need in near real-time. It’s even possible to envisage a future where the in-memory solution entirely replaces SAP BI and other SAP systems.

The quest for better BI

This potential BI revolution, however, is likely to be years away. SAP’s first steps have been to offer some out-of-the-box applications which will connect into standard SAP ERP to fill perceived gaps, such as an enhanced succession planning application as demonstrated at the recent Sapphire conference.

The other initial target for HANA implementations will be to connect the appliance to existing SAP BI systems performing a similar role to the existing SAP BI Accelerator solution, but providing direct reporting via SAP Business Objects. Where this leaves the future for the BI Accelerator is unclear, but it is likely that HANA will replace it in the near future.

The main aim of both the HANA and BIA appliances is to achieve a marked improvement in reporting speeds. Performance is, of course, one of the essential pillars of good business intelligence and HANA certainly heralds the potential to achieve unprecedented levels in this aspect of business information management.

What HANA does not address, however, is the provision and preparation of the data which it consumes from the SAP BI system. It is common that business complexities require data to be extracted, transformed and enhanced, then loaded into a data warehouse such as SAP BI so that it is fit for users to employ in their reports.

Best practise for any reporting system dictates the use of such an ETL layer to cleanse and homogenise data from multiple sources to provide a “single point of truth”; effectively a dataset on which business users can depend to give them reporting consistency. Yet achieving this undoubtedly requires utilising the standard SAP BI toolset. For all the advancement that HANA brings, what is clear is that it still relies on a solid foundation in terms of data provision from SAP BI.

Business-ready

In our experience, the majority of data models in SAP BI are not standard, which means that essential ETL work must take place before end-users can utilise the data. In future, SAP may make further developments which will allow HANA to deal with these issues, such as using SAP Business Objects Data Services to provide real-time ETL and potentially remove the need for this data warehousing layer entirely. In the meantime, however, enterprises may take the plunge with HANA, but not get the business intelligence ‘silver bullet’ they are looking for because of weaknesses in the underlying data models in SAP BI.

It’s clear therefore that investing in HANA now may not solve the more immediate BI performance pains that many organisations face. In the current climate of austerity, a more cost-effective approach may be to first make the most of existing BI capabilities; as a consultancy we have seen many instances where much can be done to maximise existing assets.

For many companies, we believe the smartest strategy is therefore to optimise performance with current systems and address existing pain points. When this performance optimisation is achieved, it will also make the decision and business case for a future investment in HANA much more straightforward.

Achieving better results

The fastest way to optimise SAP BI is to conduct a performance review investigation to identify and address problems which may have arisen both at the front and back end of the system.

Front-end performance problems usually relate to speed; a user will run a query, but experience long wait times for the results to be retrieved. Factors which may affect front-end performance might include issues such as network speed, local PC configuration, or even simply the compatibility of the various software components or combinations being employed. Additionally, users selecting large volumes of data or perhaps that the database design of the underlying tables is not optimised, can also have a detrimental effect on performance.

Back-end issues generally relate to long data processing times during the extraction, transformation and loading (ETL) stage. For instance when data is uploaded from a source system, inefficient transformation algorithms can result in an excessive amount of time for extractions to complete, hence the provision of up-to-date business data is delayed.

Discovering these problems requires a business-driven approach which should involve thorough consultation with the business to establish what problems are being experienced, and what they believe may be the root cause.

Now or never?

Even in its current format, many organisations will undoubtedly see the benefit of investing in HANA. Yet, until it reaches full maturity, enhancing their existing SAP Business Intelligence could be a faster, more cost-effective route to achieving timely, accurate, and credible data and so improving ROI on existing infrastructure.

However it is achieved, the ability to access this ‘single version of the truth’ will allow business decisions to take place based on the clearest and most accurate reflection of current business performance.