Musical chairs with the hyperscalers
Blog: Capgemini CTO Blog
Over the last two years I’ve been a strong advocate that the road ahead for the SAP S/4HANA ERP play is keeping a standard simplified core with differentiation and extensions using cloud platforms. My tryst with customers in day-to-day successful meetings reinforces this belief, and so when I see SAP’s continued romance with the hyperscalers, it’s positively encouraging, but constantly changing, the announcement of the industry clouds by SAP may cause another rethink
The story is simple and powerful – Integrating technologies to make business processes and applications more robust, agile, and flexible. But, now comes the tuff part when it goes beyond a single use case. Should I use IoT with Leonardo on Microsoft Azure, or, Google’s AI with my existing SAP system, or Alexa for better customer experience? Should I recommend clients towards a SaaS cloud architecture using one or two, or go for a best of breed multiple vendors? Choices and messages sometimes confuse and conflict.
This is an area where I think consensus has not yet been reached. Hyperscalers obviously would want system integrators to commit to their platform for the majority and hence investments flow in forms of an Embrace (SAP-Microsoft) or Extend (SAP-AWS) program. But again, the option of going down a total best-of-breed route, in theory, with the API based approach technology should not be an issue at all.
As a responsible system integrator and trusted partner to several customers, these are my preliminary thoughts as SAP continues to push a strategy to own the application stack in tandem with hyperscalers.
The pros as I see them
- SAP with a dominant hyperscaler makes commercial sense as its less complex to manage.
- There are more cloud hosting options (i.e., locations) available through hyperscalers.
- SAP SCP components are developed on hyperscaler platforms and therefore are optimized.
- Hyperscalers provide a full suite of agile-supporting tooling technologies.
- Small BoB players often are acquired by the monoliths, so users automatically avoid change.
- SIs develop IP on hyperscaper platforms so rapid innovation to capture the most of SAP S/4HANA.
- SLAs are stronger and way better to avoid reputation damage of failure.
And the cons
- Without a proper approach a vendor lock in and a lack of competition is high.
- Prices may be higher.
- Some clients view large hyperscalers as competition and don’t want to work with them.
- We may miss out on innovation by not looking at other options.
- Getting hyperscaler, SI, SAP, account teams to have a single message is hard.
Evidence appears so far that, clients tend to favor a one hyperscaler play, only using the other SaaS services during unique case-to-case special needs or if prices are prohibitive. Given the newness it’s not uncommon or strange. I’m taking a more middle path thinking there is not one single path for how SAP customers will move their applications and workloads to the cloud. This isn’t a binary decision tree approach but something long term with implications that don’t get you locked in. Data lakes are better with one, enterprise application impact perhaps better with another, AI/ML are almost synonyms with the third, pricing with a next and industry-based capabilities automatically shows the logo in your head. May be the answer is this fast changing environment, is not to be 100% locked into one path and to make sure at all points you have alternatives! If you have another view? Or would like to discuss more, do not hesitate to contact me directly.