CRM on Hana a few announcement observations
Earlier this week SAP announced at SAPPHIRE Madrid that SAP CRM would be on put on HANA as part of a larger application offering called SAP 360 Customer. This solution would surround the existing SAP CRM and sales on demand solutions. I was not at SAPPHIRE Madrid, but was able to get a sneak peak of the SAP CRM on HANA and like all of you went through the publicly made materials on the solution. That being said I feel that most analysts don’t fully get ramifications of what this means when it comes to current on-premise SAP CRM soslution. Even though SAP can make statements like X amount faster it means nothing if we don’t understand the bottlenecks of the current SAP CRM system and how HANA can turbocharge the system.
SAP CRM: A search based system
For those of you who don’t work with an actual SAP CRM product on a daily basis as either a user of the system, tech support/consultant or developer may not realize that user interface requires you to either create or search for data as your first step in dealing with any type of data. We can divide these searches into two primary types: master data and business transactions. I’m not going to go into the full data model, as one could write part of a book on that topic 😉 but let’s take a quick look at the difference between the two. Master data in CRM terminology is generally the same as ERP Master Data. CRM Master Data includes data such as customers, prospects, contact persons and employees which in SAP CRM terminology are business partners. In addition goods or services that may be sold are known as products. Business Transactions describe an interaction between business partners where a communication or exchange of goods or services takes place. Common examples are customer visits(activities), customer complaints, sales orders, service requests, etc. In CRM we have two primary ways of looking up data. We can take a “customer centric” approach or “interaction centric” approach. The customer centric approach involves finding the overview screen of an account and then looking at the related data. The interaction centric approach involves looking for a particular type of interaction in the system for a broad range of criteria.
As you can tell from either approach we must search for data before it is brought up. In the customer centric approach we bring up a business partner and then several “searches” are executed in the background that show the related data. This related data includes transactions that are related to that business partner and in addition other partners that are related. For an interaction centric approach we bring up a search screen for a particular type of interaction(sales orders, complaints, leads, etc) and then provide criteria that may cross one or many business partners. In both cases however we are always executing the same technical search code in the background the difference is that on part of the UI the search has been automatically limited to a single business partner where in the interaction centric approach we not limited to a single business partner unless we choose this. In both cases the searches executed end up being part of the day-to-day usage of the business users of a SAP CRM system.
Business Transction Search Performance Problems
The business transaction search can be very slow if there were no optimizations to made to it. The reason being is that business users rightfully demand the ability to search on almost any attribute of a business transaction. As we understand that if we try to perform searches on a OLTP data model we always encounter tradeoffs when search. The code for searching for business transaction in SAP CRM is quite complex and can make your head spin when you look at it for the first time. Some of the early attempts to optimize the search involve built-in tables within CRM to index the data for searches. The most infamous table CRMD_ORDER_INDEX is a table that adds no business value beyond making transaction search feasible in SAP CRM.
Speed up Search, Speed up SAP CRM
As SAP has realized that search is a key bottleneck in SAP CRM attempts have been made to speed it up. With introduction of Enterprise Search tools using the TREX SAP has been able to offer ways to speed up search by adding on additional layers. However the drawback of these approaches is that data must still be indexed and requies additional landscapes to manage. This is where the promise of of SAP CRM on HANA comes in play. For the first time in the history of SAP CRM we can actually remove layers to gain performance optimization. By using in-memory computing to power the search on the same database our searches have the promise of not causing the “wheel of death” on the SAP CRM system to be shown. That promise of speeding up a fundamental use case within SAP CRM why I see the true value of putting SAP CRM on HANA. Once we no longer have to worry about performance there are so many other creative things we can dream up such as “real-time segmentation of our customers” instead of waiting for sales data to be available and a batch job to run to put our customers in the correct market segment.
Wait aren’t you some salesperson, or getting paid for this?
As I said earlier beyond getting a sneak peak, I have no paid involvement with promoting this. My opinions reflect the materials made available to me about the solution and my experience working with a current SAP CRM 7.0 on standard database from a technical support perspective. The wheel of death is the number one complaint of my CRM users and any solution that promises to eliminate that grabs my attention. As the old saying goes the “devil is in the details”, it will be interesting to see the cost, time, and effort to migrate existing SAP CRM customers to the platform. If the cost and effort are reasonable and can provide the expected ROI, I would expect the general excitement and adoption to be faster than the webclient UI. Yes we can always throw more hardware at a SAP CRM system to make it faster, but I think this is case where instead we are trying to treat the cause of the problems instead of the symptoms. I think the real debate will be on ROI and effort and not on the fact that performance will increase. That being said this is not recommendation to drop what you are doing and go install this immediately. Rather I feel that all SAP CRM Customers must take a hard look at this solution and can not look the other way anymore when it comes to HANA.