I do like Barry Manilow’s Grammy winning song, Cobacabana and every time I hear CO-PA (Controlling Profitability Analysis) I hear that song. Even more so now that I am hearing about the CO-PA HANA solution – it closely rhymes with Copacabana.
Last week ASUG had a great webcast on High Performance Profitability Analysis with HANA CO-PA. Carsten Hilker is the SAP Solution Manager for LOB Finance and this solution and was the primary presenter during this webcast.
As usual the legal disclaimer applies as solution is not available yet so everything is subject to change. SAP is aiming for a Q4 release this year.
Agenda:
High performance profitability
HANA
HANA CO-PA Accelerator
Implementation Considerations
Business Transformation Services
Demo
High performance profitability
Figure 1 – Source, SAP
CO-PA has been around for several years, since R/2, and deals with large amounts of multi dimensional analysis data. The technology constraints are now gone and we are now free to do what we want and process massive amounts of data. “In memory databases have arrived” according to Carsten. All the data is at your hands.
SAP HANA
A HANA background was given; you want immediate answers. We have been following 30 year old design principles, dependent on hard drives. We have been designing around inefficiencies of hard drives. As a result, we have a complex BI landscape today.
Figure 2, Source: SAP
We have a complex landscape today as shown in Figure 2. To solve this they looked at running in main memory. This has been covered in previous blogs. A column-oriented database will have everything covered separately. Today we only look at a few fields with Profitability Analysis but HANA will give you the ability to look at more data.
Figure 3 BI with HANA, Source: SAP
Figure 3 shows that we are operating with the lowest level of information we need, like our transactional ECC system.
Figure 4, Source SAP
Figure 4 lists the customers using HANA.
Figure 5, Source: SAP
Figure 5 shows HANA being used as an application platform. SAP is creating content with HANA based on feedback from customers. HANA CO-PA is the number one application that customers are using, to be used by both sales and finance.
HANA CO-PA Accelerator
IDC said “improved reporting drives better decisions”. Today limitations are significant as we cannot get to the data fast enough. We have a dependency on IT today as they control how much data you have access to and the levels.
Figure 6, Source: SAP McKinsey’s Big Data
We can now process data, enrich it with data (such as CO-PA data with sales order data). Big data can increase operating margin by 60%, per McKinsey.
Figure 7, How CO-PA looks today, Source: SAP
Figure 7 shows the traditional ERP stack, and CO-PA gets information from CO and SD.
Figure 8, CO-PA Today using Report Writer, Source: SAP
Figure 8 shows the old standby Report Writer, which has been around a long time.
Figure 9, CO-PA with BW, Source: SAP
Figure 9 shows data being transferred from ERP to BW with different data with different levels.
Figure 10, Source: SAP Summary of Challenges
Carsten reviewed the summary of challenges with the current CO-PA situation today as shown in Figure 10.
Solution is that we take data from traditional database and replicate to HANA. Once loaded, it will allow archived data to be loaded to HANA.
Figure 11, Solution: Source: SAP
Figure 11 shows archived data being loaded to HANA, giving you the complete dataset of everything that has been posted, including archived data.
Figure 12, Improving Performance of CO-PA Allocations, Source: SAP
Figure 12 shows allocations from CO to CO-PA will run 50% faster as we are going to HANA as the read performance is improved.
Figure 13, Report Writer runs Faster with HANA
Report Writer can read information from HANA and run faster as shown in Figure 13
Figure 14, Ability to run CO-PA with BI 4.0 Tools
BusinessObjects Explorer, BusinessObjects Analysis Office can all be used against CO-PA HANA with all dimensions available as shown in Figure 14. HANA will have line-item detail and aggregations are no longer a bottleneck.
Figure 15, Reporting Example
Figure 15 shows a reporting example with 120 million records that processed in less than one second.
Figure 16, Allocation Example, Source: SAP
Carsten then reviewed an allocation example where it took only 30 seconds to read the reference data.
Implementation Considerations
Figure 17, CO-PA HANA Implementation
Carsten said the CO-PA HANA implementation is comprised of consumable components, with ERP CO-PA ABAP add-ons to allow reporting from a connectivity perspective. It has a CO-PA HANA model accelerator. A program that reads the operating concern from CO-PA in ERP and creates an XML file with definition of operating concern and HANA will read that to create the data model.
Installation is via a support package/note. Implementation does not have to take 3 months, according to Carsten.
Figure 18, Data Replication, Source: SAP
A number of implementation scenarios are envisioned. If you want HANA based reports you need to bring over line item and master data. For this recommend use SLT (system landscape transformation) that allows both master data and transactions. SLT is included in HANA depending on package you buy.
If you want accelerated ERP reporting you do not need master data replication and use the HANA CO-PA accelerator and SLT is not required.
Carsten says replication is real-time. Is that needed? More than once a month is needed. Closer to month-end you want to know what the profitability is.
HANA CO-PA Accelerator is free for ERP/HANA customers. Implementation Services offered will be the Rapid Deployment Solution including documentation, installation and business transformation workshop.
Figure 19 Implementation Components, Source: SAP
Figure 19 shows that you need hardware, and the SAP HANA standard or extended version (differentiated by the inclusion of SLT or not), the Accelerator which is free and the implementation services.
Figure 20, Software Requirements, Source: SAP
Carsten provided the information in Figure 20 so you can determine if your software requirements situation fits CO-PA HANA Accelerator.
Figure 21, Source: SAP
Business transformation gives you instant access to all the data you already have. Carsten said maybe there are fields you can use now and you can change allocations and look at your archiving strategy. You can now consider doing shopping cart analysis – who bought what and when? SAP offers HANA CO-PA workshops and provides you with an assessment.
Demo
How do we get to the data? The demo was with BusinessObjects Explorer, “google like” search and hands on ad hoc analysis.
Figure 22 – BusinessObjects Explorer Demo, Source: SAP
Figure 22 shows the front end with the CO-PA HANA data. The left hand side is the measures which are the key figures you have. The facets are on the right on how to select. Lower half includes the visualizations. The top right corner shows an active counter of records and will be continued to be filtered and how many records and how long it took.
Figure 23 – 3.2 Billion Records, Source: SAP
The response time with 3.2 billion was just as fast as going against 10 million records.
Question & Answer
Q: Financial closing and consolidation and consolidation is on BW; how will that work?
A: Financial close depends on company; in Europe CO-PA is part of the close
This solution is limited to CO-PA
Q: We do heavy CO-PA allocation on BW now and use realignment of master data
How will realignment work on reference data stored?
A: Maybe reason you brought CO-PA to BW is not there anymore. Consider bringing allocations back to ECC.
Q: CO-PA data in BW is supplemented with sales data and not captured in core CO-PA information. Will that work with HANA?
A: Carsten said this is an enrichment scenario and a join in HANA and reporting scenario
Q: Will HANA work with BPC?
A: It is separate, and now CO-PA is in HANA and bring it in to PCM and solutions will grow together.
Q: BPC on HANA, how will data be stored together?
A: BPC will come in 2 different flavors – BPC Netweaver on HANA
HANA is intended to be a multi-tenant environment, BW, BPC and operational data stores. HANA CO-PA will be in the operational data store area. The longer term plan – 2012/2013 – version of BPC that works natively on HANA without BW (similar to BPC on SQL Server today per Carsten.
Summary
This is an very interesting business use case for HANA and the possibilities. It is nice to see that HANA will provide performance improvements to old tools such as Report Writer. With the CO-PA HANA Accelerator is it interesting that you do not have to create the data models from scratch either; it is like providing BW content but customized to your own particular situation.
Upcoming HANA Webinars/Sessions This Week:
Wednesday, August 10, the SAP Customer Solution Adoption team will provide a Webinar on SAP HANA Operations. Click here for more information.
Thursday, August 11, ASUG is having a webcast SAP High-Performance Analytic Appliance (SAP HANA) Software - What You Need to Know to Implement SAP HANA for Your Business - Register here (you must be an ASUG Member to register).
Friday, August 12, SAP Mentor John Appleby will provide "an overview of the HANA product, architecture and integration. Tips for project success and finding use cases for SAP in-Memory technologies" at SAP Inside Track Newtown Square, hashtag #SAPITNSQ. Register here.
SAP Inside Track is a free event and open to all, but registration is limited so register soon.
