Part 2 – Scaling and Flexing: Data Vault BenefitsAugust 3, 2020
Breaking the barriers – Getting your SAP data on any cloud data platformSeptember 7, 2021
Enterprises running SAP Business Warehouse (SAP BW, BW/4HANA and BW on HANA) are keeping a close eye on challengers like Snowflake. Is cloud data warehousing the answer to all the challenges for organisations with a large SAP footprint? And how does a cloud data warehouse fit into the data architecture? Should it replace SAP BW or is there still value in SAP BW? In this blog post, Jan van Ansem wraps up a discussion about this topic that has been unfolding on LinkedIn over the last few days.
Why do most enterprises with a large SAP footprint run SAP BW?
The dominant position of SAP BW is easy to explain from a historic perspective, but it would not do justice to SAP BW to ignore its current strengths as well. Let us start with the latter. SAP BW is still the only data warehouse platform which delivers all data warehouse features in one platform.
It supports data modelling, ETL, data streaming, master data management and other features all in one place. It offers first class out-of-the-box functionality for many challenging areas in data warehousing such as slow-moving dimensions, inventory reporting, multi-language master data, currency translation and hierarchies. SAP BW also strongly promotes consistency in design and development and has a mature authorisation framework. These make governance and data security relatively easy to manage. If you consider moving to another platform, you will need to think about all of these and come up with a solution yourself to ensure you do not throw the baby out with the bath water. All these features can be replicated to Snowflake and other platforms, but it will take time and discipline to do so.
The above explains why SAP BW is still popular, when you then take the historical context into consideration it is easy to understand why it is so dominant. SAP BW has been around for well over 20 years, with continuous innovation and proven reliability. For a long time, SAP BW was the only tool which could get data out of an SAP source system efficiently. SAP BW comes with heaps of business content, which has helped the global adoption of SAP BW significantly.
Why would organisations look at alternatives?
In my 20+ years of working predominantly with SAP data warehousing technology, I see the following challenges time and time again:
- Struggle to integrate SAP and non-SAP data at scale in a (cost) effective and meaningful way.
- Unable to monitor and manage the true total costs of ownership (TCO) of the data platform.
- Lack of synergy between data & analytics teams: The SAP data team and other data teams do not speak the same language.
- Long time to market for change requests.
- Inability to support business driven innovation projects (fail early, fail fast and fail often)
SAP has continuously brought innovations to SAP BW to alleviate these problems. Smart Data Access & Integration, Field based modelling and virtual modelling go a long way to make SAP BW more agile and, in some ways, making data more accessible from outside the SAP platform.
I often see that these innovations are either not fully understood, or not available on the specific BW version the customer is on, or the costs associated with some make it prohibitive to use it. Not all problems have a technical root cause either. The long time to market is often at least partly caused by over-complicated governance and change processes. The poor ability to manage TCO effectively is partly down to SAP’s murky licensing model.
Is Snowflake really the silver bullet to resolve all these problems?
Of course not. Your analytics capability depends on a lot more than technology alone (governance, processes, culture, strategy). Technology can be a blocker or an enabler when pursuing certain goals.
Today, those goals for analytics are usually formulated along the lines of becoming more result oriented (evidence-based decisions, relying on metrics and measures), ensuring business functions can use analytics tools and methodologies independently and optimizing business processes with technology and data (AI). To achieve all that, you do need technology that makes data easily accessible and easy to integrate with whatever tools the business are using. Here is why Snowflake is so very well suited to fulfil this role:
- One copy of data but many (independent) workloads
- Near unlimited performance and scale
- Zero-copy cloning to support testing (in production) and innovation projects
- Near zero maintenance – as a service (No more SAP notes, upgrades, server patching!)
- Open, easy 3rd party integration (no need to check the small print in your contracts)
- Use the best of breed and most popular tools in ‘data’, instead of proprietary products and technology.
Snowflake is not a silver bullet, but certainly has attractive and unique features which will help organisations to become more data driven and agile.
Will SAP BW live to see another day?
Certainly. Some organisations will be happy to stay on SAP BW without moving to any cloud data warehouse any time soon. Analytics will mainly be based on SAP data, but perhaps that is enough to satisfy the business needs. Other companies will utilize a cloud data warehouse as an innovation platform or for non-SAP data, and decide to keep the SAP data in SAP BW.
Even those organisations which do decide to move to Snowflake will not get rid of SAP BW in a hurry – you cannot replace a mature data warehouse in a day. In fact, 9-12 months would be very impressive in most companies.
My advice? Listen to your stakeholders, their current pain points, and their requirements for the foreseeable future. Based on that, decide which technologies will be the best fit for your organisation. Do not forget to look at what else you might need to change; the success of your analytics capability is not just about technology. Snap Analytics can help you with this journey.
When making a choice make sure you understand the strengths and the weaknesses of the different technology platforms – and leave your biases at the door.