99.7% SLA in SAP RISE: What 26 Hours of Downtime Means for Your ERP

99.7% availability. At first glance, a respectable number. In practice it means: up to 26 hours of unplanned downtime per year. For your business-critical ERP system. For the platform that runs your order management, your finance and your logistics.

The standard SLA for SAP RISE productive systems sits at exactly that figure. SAP raised the availability guarantee from 99.5% to 99.7% in 2021, a step SAP communicated as "raising the bar" (SAP News Center, August 2021). In absolute terms: the guaranteed maximum downtime dropped from 43.8 to 26.3 hours per year. Progress, but the question is which benchmark you measure it against.

Because the infrastructure SAP RISE runs on delivers significantly more. AWS guarantees 99.99% for multi-AZ deployments, which translates into a maximum of 52 minutes downtime per year. Azure quotes 99.95% for Availability Sets, around 4.4 hours. SAP uses that infrastructure but only passes 99.7% through to the customer. The gap between what the platform can technically deliver and what SAP contractually commits to is meaningful.

Industry standard or outlier?

Gartner positioned this clearly in May 2025. In a report covered by The Register on 8 May 2025: users find RISE service levels below industry standard. This is not an isolated opinion, but a systematic observation from customer reality.

An upgrade to 99.9% availability is possible with SAP in principle. Downtime drops to a maximum of 8.76 hours per year. The price is significant: Gartner quotes a 30 to 50 percent uplift on the software licence. Jan Cook of Gartner puts it directly to The Register in May 2025: "We see very few examples of customers buying that."

The implication: the vast majority of RISE customers run on the standard SLA. 26 possible hours of downtime per year are not the exception. They are the norm in the RISE landscape.

What makes the situation harder: the SAP SLA defines incident response times but no binding resolution time. SAP commits to react to an issue within a defined window. When the issue is actually resolved is not contractually guaranteed. For a business-critical ERP system that is a relevant gap.

Add to that the question of measurement. Planned maintenance windows are excluded from the SLA calculation in many contracts. What SAP classifies as "planned maintenance" and how it affects effective availability is hard to verify without your own measurement data.

What you can do: five concrete steps

1. Set up your own availability measurement

Do not rely solely on SAP availability reports. Implement independent monitoring that captures the actual reachability and response times of your RISE systems. Only with your own data can you confirm that SAP delivers the committed availability, and only with your own data do you have a defensible basis for conversations with SAP.

This does not require complex infrastructure. Simple availability checks at defined intervals, logged cleanly and evaluated over time, are enough as a starting point.

2. Read the SLA definitions in the contract closely

Not every minute of outage counts the same. Check how your contract defines "availability": which maintenance windows are excluded? How are partial outages handled, where the system is technically reachable but reacts so slowly that productive work is impaired? Is the SLA different for productive, QA and development systems?

These details determine what 99.7% actually means in your specific case. The headline number says little if the calculation basis is not transparent.

3. Calculate the cost and benefit of the 99.9% upgrade

A 30 to 50 percent uplift on the software licence is a substantial amount. Whether it pays off depends on your specific business case: what does one hour of downtime cost in your organisation? How many users are affected? Which business processes stand still?

In many cases the math will show that the upgrade is not economical, at least not at current pricing. But that conclusion should be reached deliberately, not by default because the option is unknown.

4. Use SLA performance as a renewal lever

If you have collected your own availability data over twelve months or more, that data is a concrete negotiation asset. If actual availability exceeded 99.7%, it is an argument for SAP to lift the SLA without an uplift. If it came in below, it is an argument for renegotiation or compensation.

Without measurement data there is no negotiation position. With measurement data an abstract SLA number becomes a concrete conversation.

5. Separate response time from resolution time

Make sure your internal stakeholders understand the difference between incident response time and time to resolution. SAP guarantees the former, not the latter. If your CIO or business unit assumes a critical incident is resolved within four hours because the contract states "4h response time", that is a misunderstanding which becomes expensive in the moment of truth.

Communicate internally what the SLA covers and what it does not. That sets realistic expectations and reduces escalations built on incorrect assumptions.

Conclusion

99.7% availability is the standard in SAP RISE contracts, and for most customers that will not change in the short term. The 99.9% upgrade is priced in a way that makes it uneconomical for a large share of customers.

The relevant question is therefore not whether 99.7% is sufficient, but whether you know what the number means in your case, whether you measure if SAP holds to it, and whether you use the results at the next renewal as a basis for negotiation. The SLA value in the contract is a starting point. Whether it works for you depends on what you do with it.

Sources:
Gartner via The Register (8 May 2025): "Users find RISE with SAP service levels below industry standard."
Jan Cook (Gartner) via The Register (22 May 2025): "30-50 percent premium." / "We see very few examples of customers buying that."
SAP News Center (2 August 2021): Availability SLA raised from 99.5% to 99.7%.