Support for the Cloud World – Opportunities and Considerations
In this post, I’ll describe why and how support in the cloud world can be fundamentally better than traditional support.
1) What is different today?
Business applications, including business critical applications, are more and more moving to the cloud. Looking into a support case in on-premise business, support engineers typically have access to the customer’s product information, the release, and the software package. To understand the issue the engineer needs a detailed description, often access to the customer’s system, plus a detailed verbal description by the end user.
In the cloud world support can be fundamentally different. The cloud provider, which is often the same company that provides the support, has a lot more information: In addition to the exact release and software package the customer is running on, the last release change date, anonymized usage data, customer configuration data etc. To avoid any misunderstandings, of course customer business data that is owned by the customer and especially personal data, must not be accessed or leveraged for this purpose in any way. At the same time, artificial intelligence (AI) algorithms get much more powerful and are nowadays able to capture and draw conclusions from multi-dimensional inputs that are ideally suited for the support world. All this needs a direct connection between the cloud operations (data centers) and the support organization.
2) What does this mean for support?
A key challenge in support is to provide the exact support answers and content that are relevant to the customer’s specific issues. We have seen that more than 80% of customer cases are related to known issues. At the same time, SAP has hundreds of thousands of knowledge repository items relevant for support. If we also include threads from communities, this figure is even significantly higher. Therefore, all known customer data needs to be leveraged to provide relevant content specific to the customer’s needs.
Furthermore, especially in the support context, the cloud deployment model opens up a completely new way how processes can be executed. Customer system logons are much easier (of course we still need the customer’s permission for a support engineer to log on to the customer system). However, because the provider now has so much more information available, the whole support model can be put upside down: the support provider does not need to wait for the customer to contact them for an inquiry. Instead, based on the availability of a much broader data base, the cloud support provider can proactively give the customer user the right information before they run into an issue which would typically make them reach out to the support provider. Prevention instead of escalation would be the new ideal support scenario.
In addition, cloud systems provide a more standardized view. For example, there are not as many versions, typically less configuration options, and the variants between hardware and software are also reduced. This standardization also means less “issue clusters” and “solution clusters” – meaning more than one customer likely faces the same issue. Even though this increases the possible impact on many customers, it also increases the opportunity to solve cases faster and in a more automated way thanks to the standardization. It also opens the door one more time to preventive support.
So the targeted recommendations to customers can take into account generic information like hot or trending content from the support provider in addition to customer-specific information. Both sides need to be leveraged optimally. Customer-specific information derived from click behavior and search terms being used, etc. can be considered to improve the customer experience towards self-service scenarios or targeted predictive notifications. It’s also a value-add to look at similar cases that other customers ran into and which recommendations helped them. The increasing ability to “track a process” also enables simulations and better testing of new releases. These indirectly support the prevention of issues.
You may now ask why the support provider even needs to reach out to the customer and why they can’t just fix the issue themselves (here: self-healing software). Of course, that is the intention as well. But in the majority of cases customers reach out to the ERP support provider because, at least for them, something is unclear. So, it’s often less about fixing something in the coding than fixing something in the user’s understanding (and often, at the same time, a hint to the software provider that something in usability, context documentation, etc. should be improved).
3) What are the opportunities?
For the support provider and the support engineers, support for cloud-based applications needs to have a much stronger focus on knowledge-based support which includes SAP Knowledge Base Articles, product documentation, how-to guides, threads in communities, and so on.
The support engineer’s role will change as well. SAP’s predictive and preventative support features and functionalities help resolve customer issues before they even occur or answer customer questions in real-time. So, the support engineer will become a knowledge and digital-guidance expert and move away from processing one customer ticket after another.
The software provider has to focus on the overall infrastructure and processes, such as real-time customer-tailored interventions – including contextual relevant customer- and user-specific notifications – that help reduce the volume of cases and achieve higher customer satisfaction through faster responses, better self-service quality and better knowledge quality. This focus requires significant investments into AI based on internal services and in working on new customer-facing, simplified digital customer touchpoints. This could include developing case-matching services so that cases issued by different customers with the same root cause can be identified which will help reduce redundant effort by the support team. As a result, the standardization could lead to better automation and, in the end, faster case-resolution times.
Data protection and data privacy topics also become even more important. The level of granularity on which data is collected and the purpose for which it is used need to comply with all legal and ethical standards. AI will help as well to identify knowledge gaps and support creating new knowledge.
In the cloud world software providers can leverage significantly more data sources than in the on-prem deployment model. At the same time AI is making tremendous progress to draw conclusions from multi-dimensional data input.
This together allows predictive and preventative support providing better self-service opportunities for customers and faster responses to customer cases submitted. So with the cloud deployment model a much better overall customer experience can be provided.