Mr. Behrens, it’s almost a year now since SAP surprised the IT world with this revolutionary in-memory technology SAP HANA. What is, in your opinion, so revolutionary about it?
Holger Behrens: From our point of view, the revolutionary thing about it can be summed up in one sentence: SAP HANA now enables our Sprint customers to turbo-charge their businesses. It enables them to analyze data in real time and spontaneously again. This was hardly possible any more with many business analyses, due to the greatly increased quantities of data. The response times – for example, to fill OLAP cubes for evaluation – have increased significantly. Such analyses are far from real-time. But that should now change with SAP HANA. The buzzword among data analysts is data surfing. Let me explain: In a business warehouse, you only move around in the data that exists there – and that must be synchronized constantly, too. But with SAP HANA, you surf in the raw data, in real time, and at an incredible speed.
So that means more RAM, so more costs, doesn’t it? Higher transaction speeds and shorter response times at any price – doesn’t that exceed the budget of most projects for small businesses and midsize companies?
Holger Behrens: On the one hand, SAP HANA is a storage technology, on the other hand it’s technology for optimized data storage and data analysis. It’s not simply about an excessive upgrade of the RAM data storage. The unmistakable advantage of this new in-memory technology is its ability to process extremely large quantities of data in real time.
One of our pilot customers in the B2B sector took quite a pragmatic approach when it opted for SAP HANA: The costs would have been similar for a business warehouse installation and the subsequent design of the analysis cubes as for the SAP HANA project, which provides many more opportunities in the medium and long term. I think that’ll be the motivation for many future SAP HANA projects – that of not investing in a “conventional” business warehouse when in-memory technology already includes BW capabilities.
Small businesses and midsize companies in particular expect calculable costs. Is that possible with SAP HANA?
Holger Behrens: Our philosophy of fixed-price quotations for implementation projects and SAP add-ons is tailored to the midmarket, and we will be sticking to it for SAP HANA, as well. Customers now benefit from our experience with data analysis in real time, as well as from our in-depth industry knowledge from our Sprint product line with FOODsprint, TRADEsprint, and so on. We work closely with SAP and are directly involved with the first implementations in the midmarket. This means we have already gathered some SAP HANA experience for midsize companies and we can bring this with us to customer projects at a later date. Based on implementation experience and the preconfigured SAP HANA Edge edition solutions, we will put together midmarket packages for a fixed price.
How will you make it easier for companies to migrate from SAP Business Information Warehouse (SAP BW) to SAP HANA?
Holger Behrens: SAP itself has announced preconfigured rapid deployment solutions (RDS) for SAP BW to SAP HANA, including implementation work. This, of course, simplifies the switch from an existing SAP BW solution to SAP HANA. Certain reports are already preconceived and configured. We’re planning fixed-price quotations here, too. However, we need to complete a few more projects before we can create realistic quotations based on them.
Does that mean that the data warehouse is dead?
Holger Behrens: I wouldn’t put it quite as drastically as that. There’ll be a significant shift on the business warehouse market in favor of SAP HANA. Especially with the SAP HANA Edge package, which SAP announced recently and which is aimed at the midmarket, we hope to be able to convince a good number of customers to opt for the new technology.
What added value do your midsize customers get from SAP HANA now?
Holger Behrens: The benefit is two-fold: First, there’s SAP HANA’s real-time component, and second, there’s the variability of the analyses. Ultimately, you access the ERP core data in real time for your analyses. This means production and distribution can be aligned more closely with market needs at the time. Response times at the computer are greatly accelerated. For example, with a conventional data warehouse, a customer can display a sales overview only after a few minutes – that’s how long it takes the system to analyze and process the data it contains. With SAP HANA, the whole thing takes just two or three seconds. And sales managers are of course happy about this, because they don’t have to wait around for minutes for their analyses to appear. In other areas, the time difference between “classic” OLAP analysis and in-memory technology is even more dramatic.
Can you describe such a scenario?
Holger Behrens: Here’s one example: We have various customers in the food industry who regularly receive POS data, that is, information from the point of sale. But what use is it if the sales information from the stores can only be analyzed when the data is next synchronized? It takes longer to react to requirements and fill up the supermarket shelves again. But, especially with seasonal business, it is important to know fast what the hits and flops are on the market. If you supply first, you have a better chance with the consumers.
And, thanks to SAP HANA, production managers at food companies, for example, can plan production not with market data that’s already a week old but with real-time information that shows them what’s required tomorrow. They can then react and adjust their production lines accordingly. This means that there are plenty of popsicles when the weather’s hot and the demand for ice cream increases. With producers of fresh goods or ice-cream manufacturers, timely analysis of data is a must for making the flexible production and distribution decisions that are needed for economic success.
And approximately how long should a midsize company expect an implementation to take?
Holger Behrens: Here, too, we cater to the needs of midsize companies and plan in weeks rather than months. Of course, we can’t predict exactly yet and it also depends on the size of the company and the volume of data. However, based on our experiences with the pilot projects, we can already estimate implementation times of between eight and 16 weeks for our midmarket customers.
And what will happen next?
Holger Behrens: The next step will be for whole processes, such as production planning or order management, to run faster with SAP HANA. That’s still a long way off though. For the time being, we’re concentrating on data analysis for reporting tasks. I think that 2012 will be an SAP HANA year for us. We already have more SAP HANA projects on the horizon.