Making sense of where SAP customers stand on cloud adoption is not a simple issue. By now, most SAP customers have experience with SaaS apps, as well as managing cloud workloads inside - and outside - of SAP.
Then we have RISE with SAP. One of SAP's core pitches with RISE is helping customers move more SAP workloads to the cloud - and simplify/manage the customer's hyperscaler relationship, at least for SAP workloads.
One open question with RISE: how will smaller SIs respond? SAP leadership has told me, on more than one occasion, that their midmarket partners are either on board with RISE, or in the process of getting up to speed. I am talking to a number of these partners now, and taking a pulse: do they position with RISE, against RISE, or take a more RISE-agnostic approach?
As you might expect, there isn't one uniform answer. One firm on my short list is Syntax. Syntax has been in the ERP services market since 1972, but they've changed with the times. A core part of their services is providing SAP customers with cloud hosting options, from public cloud to their own Syntax private cloud. Of course, that includes S/4HANA migrations.
The Syntax - ASUG annual cloud survey: surprising data points
For the last five years, Syntax has partnered with ASUG, Americas' SAP User Group, on an annual cloud survey of SAP customers (you can get the 2021 white paper, free with sign up). Note: though this annual survey is based on North American users, ASUG also issues an SAP survey with DSAG, which focuses on SAP customer cloud adoption, attitudes on RISE with SAP, and SAP S/4HANA. On diginomica, ASUG CEO wrote about the contrasting views on S/4HANA cloud adoption from US and German SAP customers.
As 2021 wound down, I caught up with Roan Low, Solutions Architect at Syntax, to talk about all of the above. But first, a couple highlights from this year's Syntax and ASUG survey:
- Primary drivers of cloud innovation include reduction in costs, improving business processes and competitive edge. There was also a 15% increase in cloud newcomers (customers who have been on the cloud for less than two years), and 34% of respondents identified as cloud "innovation laggards."
- 73% of respondents not currently using cloud services are planning to consider them. The most popular cloud deployment is the hybrid cloud at 46%, followed by the public cloud (28%) and the private cloud (25%). Only 38% of respondents use a single cloud provider; 61% are using multiple cloud providers.
So, Mr. Low, were they any surprises this year? Yes, one in particular. As he told me:
We asked customers, 'How many workloads are you running in the public cloud, private cloud, or on-premise?' All three went up. That shook me a little bit. I would expect one of them to remain stagnant or drop, and the other two to go up. What we specifically saw was the on-premise workloads went up. [Author's note: the increased workloads were: public cloud 3.7%, private cloud 3.2%, on-premise workloads 5.8%.]
Not all data points are self-explanatory:
I don't have a very good insight as to why I'm seeing that, to be honest, but the data is real. I'm not disputing it. It just really opened my eyes to the fact that while the cloud is definitely a solution of choice, or tool of choice for some companies, they are still many companies out there that absolutely need their metal.
Low believes this data shows something of a retreat from "cloud as a strategy," to "cloud as a tool ":
Cloud strategy is a term I hear a lot. I think that's given way a little bit to the view that cloud isn't a strategy. Cloud isn't a solution. Cloud is just another tool that a company, if used correctly, can gain business value out of.
And yet, we did see, over the entire survey, an 11% increase in organizations spending more than a million a year on cloud services. So we are seeing an overall cloud spending trend. Low responded:
But as I mentioned earlier, that's a shift from, 'Cloud is the next best thing; it is the solution'... Instead, it's more of a selective choice, 'How can we leverage cloud, to better do X?'
Multi-cloud usage? Perhaps - but not true multi-cloud
When we say customers are using multiple clouds, I don't see that as "true multi-cloud." To me, true multi-cloud would be the ability to have portable workloads that could easily be shifted from one cloud to another, either for price, performance, scale, or functionality. That's not where multi-cloud is today. So, when I read from the Syntax-ASUG survey that about half of those surveyed are using 2-3 clouds (47% in 2021), that doesn't mean they are easily shifting workloads between them. Low responded:
The thing we need to differentiate is that this survey goes out to ERP customers. These are not typically what I would consider cloud native companies that grew up in the cloud world, and are running cloud-native applications,
ERP is light years away from where it was... but it's still a stateful application, and stateful applications have technological restrictions that make it difficult for them to truly embrace and find value in their cloud agility... If you just take SAP, and the concept of spinning up an application server on-demand for a user task, it's not impossible - but it's pretty impractical.
Whereas, if you take a cloud native application; let's pick on video streaming. It's really easy to stand up a node that can stream 300 video streams for another range of uses, and shut it down when those uses triple. SAP, with its back-end database, and its stateful nature, doesn't lend itself towards that.
I don't think multi-cloud, in the portable way I described, is a reality for most apps. Certainly not for ERP. Low:
In the beginning, I did hear a lot about multi-cloud, 'Let's be able to shift workloads from X hosting partner to Y hosting partner, based on capacity, based on price, based on use case.' But yeah, I have not seen that.
Our customers, there are a very few that do - and it's not with SAP. It's primarily homegrown applications that are able to do that. But it's not automated. It's on-demand; it's manual. For the vast majority of our customers, they are leveraging the cloud in a way that's dedicated to certain workloads.
So for example, if they've got industrial IoT with SAP, that all runs on AWS. Active Directory, and perhaps a single sign-on, that may be running on Azure, or running on a private cloud. And those workloads do not shift back and forth across the different cloud providers.
RISE with SAP - what is the Syntax stance?
If Syntax is knee-deep in SAP cloud hosting, they have to take a position on RISE with SAP. So are they for, against, or neither? Low made a crucial point about RISE: it should be understood not just as a way to manage SAP cloud workloads, but to do it on a subscription basis:
We've taken an agnostic approach. We absolutely understand the value of RISE to many customers. I see RISE as an SAP partner-enabled approach to cloud hosting, combined with subscription licensing.
We do have customers that want to do RISE, that see the benefit in RISE, and want to go in that direction. We do not discourage them. We work with them to understand where they see the benefits, and where they see the risks, and maybe we educate them about some further benefits, or some lesser risks.
It depends on each individual situation. But we are a RISE partner; we can support a RISE implementation. To date, I have not personally done one, just to put that in. But we absolutely are talking to customers about RISE, and we can support it. So it's a totally agnostic approach.
Low believes the agnostic position is the right one, not just for RISE, but for all cloud services:
It's the same thing that we do with cloud. We offer cloud hosting, private cloud, AWS, Azure. RISE encompasses the last two: Azure and AWS. As such, we don't go into a customer and say, 'Hey, you've got to do private cloud.' We're agnostic on cloud; we're agnostic on RISE. The reason for that is: what is right for us is not always right for the customer.
The real question becomes, 'What is your long-term strategy? Why are you considering this move? Why are you even talking to us? What are the business benefits you're hoping to achieve?' If you're looking at let's say, machine learning, or maybe industrial IoT, we're probably going to start asking questions, even steer you towards AWS right now. Because that's where we see most of the business benefits for those areas. If you're doing maybe machine learning or AI, maybe then we're going to start talking to you about Azure, because that's where we see some of the benefits of Azure.
I'm not going to sit here and tell you how smaller SIs should approach RISE with SAP. That comes down to their own customers and business model. I've seen a few smaller SIs bash RISE on social media. The worst part of the bashing is not the bashing - they can bash whatever they want. To a point, it's refreshing. But in many cases, they don't seem to understand what they are bashing. Customers need informed pros and cons - including informed criticism. RISE hyperbole, whether it's positive or negative, doesn't clarify anything. I've seen too much of both, frankly.
Maybe mocking RISE gets marketing attention. Maybe being "agnostic" isn't terribly sexy. But "agnostic" puts you in the position to educate customers on their options. That's what I'd like to see even more of, both from SAP and its partners. That said, I'm due for a RISE 2022 update from SAP; I'll be speaking with more SIs on the record about RISE also. I'll say this: SAP still has a lot to prove on RISE. If SAP is confident RISE can deliver, I'd like to think they would welcome this challenge. See my piece, Is SAP RISE changing the S/4HANA business case, or not? UKISUG goes behind the numbers of their newsmaking RISE with SAP survey.
A word on the cloud adoption numbers: we are talking about a region-specific survey, with a relatively small sample size (114 in 2021, 34 of which are S/4HANA customers). So, that increase in on-premise workloads comes from a pretty modest sample. But if this indeed a broader trend, I'm not terribly surprised.
While I am an undeniable cloud advocate, I do not believe shifting ERP from on-premise to cloud hosting is much of a business case. There must be a broader transformation imperative, and a reason to modernize systems - not just move them somewhere. I have plenty of criticisms of RISE, but one thing I do like: with RISE, SAP presents a coherent view of how to move a heavily-customized system back to standard - over time. It's hard to build business cases with heavily-customized ERP.
Cloud should never be a strategy unto itself. But if your strategy is serving your customers and suppliers in the most transparent and, if you can stomach the buzzword, agile way possible, then I don't believe you can really fight cloud - as the long term direction, in most industries. Aerospace and Defense may be the most notable exception.
Take this example from the Syntax - ASUG survey: "1/3 of respondents plan to have artificial intelligence (AI), machine learning (ML), and/or automation innovations in the cloud within the next year." Good luck applying AI/ML to your older, on-premise workloads. And good luck trying to absorb new predictive and process automation functionality on older, legacy systems.
That said, customers need to set their own modernization priorities. Some large SAP customers have very sophisticated IT shops - and are more than capable of pulling data out of legacy systems where needed, without upgrading them or moving them anywhere.
One thing that really impressed me: the amount of cloud innovation the surveyed customers are investing in. They might have notable on-premise workloads, but the cloud activity is hardly just lift-and-shift. Example: 54% have cloud analytics in place, with 23% planning this in 2022. ML and automation activities in the cloud are gaining ground fast. To Low's point, a sharper understanding of the business benefits of cloud seems to be coming into focus - rather than just a PaaS/IaaS tech/workloads perspective.
I asked ASUG Research for their views on how this survey compares to prior years. ASUG Director of Research Marissa Gilbert said that moving to the cloud remains a focus for ASUG members. She pointed to the number of cloud newcomers this year (15%), and expects a similar trajectory moving forward.
Low told me that in his Syntax practice, the number of large-scale S/4HANA projects was surging towards the end of 2021 - a contrast to the pandemic slow down.
Starting in the middle of the third quarter onwards, we've seen a definite uptick... We're definitely getting a lot of good discussion, and we've had some good signings around the S/4 migration.
In terms of SAP direction, Low says all of his new S/4HANA projects are cloud: "100% of them are either private or public cloud." Low added that his colleagues in Germany and elsewhere might be seeing different numbers, but that's what he's seeing in the US.
It's too early in the year to make any categorical statements, but I hope this piece is a starting point for 2022 data and debate.