| QUESTION & ANSWER |
SAP explains embedded analytics, part 2 |
 |
By Robert Westervelt, News Editor
20 Jun 2005 | SearchSAP.com |
|
|
In part two of this two-part interview, SAP's Roman Bukary points out why SAP is separating its new embedded analytics from its xApps, despite both products being similar composite applications. Bukary, vice president of SAP's analytic applications, said embedded analytics give customers much broader capability even though they are the same as SAP's xApps technology.
|
|
|
|
|
Every xApp is a composite application but only some composite applications carry the association with the xApp name.
Roman Bukary,
vice president of SAP's analytic applications, SAP
|
|
|
|
|
|
|
|
|
|
What is the difference between xApps and embedded analytics?
Bukary: Every xApp is a composite application but only some composite applications carry the association with the xApp name. The xApps are a particularly branded type of composite application. They are branded this way to show momentum in the market. They address the pain points that the SAP portfolio may not have been able to address properly.
So they are both composite applications, but will sell at different price points?
Bukary: The price point for xApps is consistent with SAP's pricing methodology. Pricing has not been determined for analytics at this time, but they will also be consistent. As we move forward in time, everything will become composite in nature. Everything will be process oriented, drawing on SAP and non-SAP elements. Analytics are the continuation of the trend SAP started, moving everything toward composites.
How will these analytics interact with the current set of BI tools that many customers have? Will customers have to throw out any investments they made?
Bukary: There is no generic answer that I can give you to such a complex question. For example, SAS provides great data mining analysis statistical processing capabilities, and for the customers that need to do that exact level of complexity there will always be a need for a vendor like SAS.
((Content component not found.))
You will need the entire BI [business intelligence] stack right now. But it probably won't exist in its current configuration of discrete, self-contained applications.
This new world of analytics that I describe will always need a data warehouse. They will always need it because they will need the context. You will always want historical data. ETL always will be required. I've heard for decades people say that over time everything will be defined, normalized and standardized, and everything will go to Web services. We're still waiting for that.
So, you admit you will be adding some initial complexity challenges for customers if they choose to use these products.
Bukary: It adds challenges only if we don't do our job well. We add more products, more features, more functionality, and it adds complexity only if we don't do our job. The guidelines we have are to make it simple, easy and consumable by an average business person like me.
Even if SAP does not exist, the need for data warehouse construction always will exist. Every company today needs to know and wants to know where are they, where do they come from and make certain guesses about what the future looks like. These companies are out there building data warehouses. With NetWeaver, we make the problem simpler for them. We at SAP recognize that you will want to take structured data inside of data warehouse and combine it with unstructured data.
All of these composite applications are grouped into four categories?
Bukary: The grouping of SAP analytics go across every functional area of SAP. Supply chain, SRM, HR, ERP [enterprise resource planning], CRM [customer relationship management]. You pick a three-letter acronym and we've got it. We didn't just go horizontal. What we did was pick functional areas vertical and extended them there as well. I claimed that SAP analytics are built for business users by business users. They are built by application solution managers, and industry solution managers take them and made them specific for banking and other industries.
What is required of customers and what type of customer are you targeting?
Bukary: The answer is that there is very little required by a customer in order to be an appropriate analytic customer. What is required is that they have a need to merge strategy with execution. Their current way to doing things is insufficient. Senior management must buy into this; it is not a departmental decision. The thing we are trying to help them avoid is a situation where every business person has the tools except their data sets are all different and all reflect different numbers, which can cause a battle of the spreadsheets.
Page 1 | Page 2
|
|
|
|
