Recently I hosted a session about operational analytics at SAP Inside Track Oslo. I realized that many people are not aware of the possibilities they have with the embedded BW in their Business Suite systems, as well as other great things you can do with HANA beyond the common usage scenarios, so I decided to raise awareness of these low-hanging fruit with this blog post.
If you’ve got BW on HANA but the migration of your ERP or CRM system to HANA is still some time down the road, you can boost the operational processes in your ERP or CRM system right now, well before you migrate to Business Suite on HANA.
There are several ways to make the speed and extended functionality of HANA available to your transactional applications even while they still run on Oracle or DB/2 as your SAP system’s primary database. From a development and implementation perspective, these are mostly quick wins which can be implemented and go live in a matter of days. All that is required is that you have an SAP HANA somewhere in your system landscape, the most likely candidate being the HANA database of your BW, or even a stand-alone HANA appliance.
Image: Low-hanging fruit. Source: File:Jackfruit Cheena02 Asit.jpg – Wikimedia Commons
Here are some of those quick wins:
Accelerate batch jobs
By analyzing long-running batch jobs, replicating specific key tables into HANA with standard SAP tools and making those long-running ABAPs query the HANA tables instead of the local Oracle or DB/2 tables, batch jobs can frequently be sped up from hours to minutes or even seconds. This doesn’t work for every batch job, but an assessment of the jobs for which this technique works with minimal or zero coding effort can be performed in a matter of hours, and then the implementation is a matter of hours per job as well. I’ve seen batch jobs go from 36 hours to 2 minutes after such an optimization was applied.
Accelerate dialog programs
With the same technique, long-running dialog steps or programs can be accelerated as well. Typical candidates are search dialogues such as searching for a customer by name and birthday, which can take very long in non-HANA databases and can be especially problematic when an impatient customer is on the phone or standing in front of the agent. Again, we can push down response times to seconds (from minutes or even hours) with little coding effort using the side car approach with replicated key tables.
Fuzzy search and duplicate prevention
HANA supports a very easy to implement fuzzy search that works just like the fault-tolerant search on Google. If you search for a name but misspell it, the search will still retrieve the correct result. For example, you could search for “tor fransen” and the database would still come up with my customer record, “Thorsten Franz,” and a number of other partly matching records, ranked by similarity to the query string. This can be used perfectly in search screens, but also to avoid the creation of duplicate customer or master data records, and it can be implemented in a non-HANA system in a modification-free way with little effort using standard enhancement techniques.
Enhance your operational applications with analytical content from HANA or BW. Making queries, dashboards, interactive diagrams, or select key figures available to users in transactional applications can greatly improve the speed and quality of your users’ operational decision-making. Making the big picture or at least aspects of it available to transaction-processing users can boost productivity immensely. There are a number of minimal-invasive techniques to add such analytical content to transactional applications, including NetWeaver Business Client and the side panel, but also similarly-looking enhancements that don’t require the use of NWBC and work with standard SAPGUI.
It’s one of the best-kept secrets that every ERP and CRM system has a full-blown embedded BW with all the capabilities of a stand-alone BW system. When you have a HANA in your system landscape, you can use the BW capabilities of your BW system to create powerful and highly interactive BEx queries and dashboards right in your ERP system – but using super-fast data access in your HANA side car. This can be achieved by programming custom virtual InfoProviders which behave like BW InfoCubes, except that instead of storing data physically, a function module is called that performs a super-fast query against your HANA side car.
This technique is especially tasty when you combine large tables in HANA with smaller tables that reside locally in the ERP system, and make the BEx queries built on top of this available to your ERP or CRM business users – without requiring user IDs and authorizations in the separate BW system, and without requiring system trust relationships, single sign-on mechanisms, firewall entries, or any of the other things that would normally be required to make BEx queries and analytical capabilities available to users of transactional applications.
These are just a few of the low-hanging fruit you can pluck when there’s a HANA anywhere in your system landscape. This is your opportunity to bring real value to your end users and get more out of HANA, for less.