BETA
This is a BETA experience. You may opt-out by clicking here
Edit Story

AI-Powered Swarming In Customer Support

SAP

By Stefan Steinle, EVP and Head of Customer Support & Innovation, SAP SE

Customer support has relied on a traditional, one-way, escalation-based tiered model and serving the needs of millions of businesses. Over time, however, the demand for more elastic models has grown in response to tectonic shifts in technology, consumer expectations, and cloud-native usage patterns.

One such model is swarming, which focuses on collaboration and flexibility. It is important to note that swarming is not a replacement for tiered support but is instead an option adopted by support teams when the need for efficiency is high or when the cases are complex.

The DNA of Swarming

While the etymology of “swarming” brings up visuals of large colonies of bees or huge thronging crowds – there’s more to swarming: the aspect of wheels being in motion and things getting done.

Swarming in business shares conceptual similarities with swarming in nature – it is busy, active, and there is a lot of buzz. Swarming is decentralized. A good contrast is the situation room, a centralized command center for problem solving whereas swarming is a decentralized and agile approach to problem solving. Swarming is characterized by collaboration and communication across disciplines till the issue is resolved. In contrast, tiered support focuses on escalations and handovers to specialized teams.

The Evolution Towards Connectedness

Historically, swarming has been around in some form – with applications in areas such as complex adaptive systems, military strategy, and organizational operations. Toyota’s Andon system is a good example of agile swarming in lean manufacturing.

ForbesSAP BrandVoice: AI-Enhanced People Development Opens A New Path To Business Growth

In the context of customer support, and with collaboration being at the core of swarming, it is clear that there is a need for smart connections across complex landscapes, multivendor scenarios, and cross-product considerations. These dynamic connections contribute to the overall responsiveness of the network.

To Swarm or Not to Swarm

While swarming seems like an answer to all your support nightmares, there are situations where swarming might not be the ideal solution for your problems. When you have cases that could be complex, time-sensitive, dynamically evolving, or need cross-functional expertise – swarming is probably the best way to go.

ForbesSAP BrandVoice: Addressing The Fear Of Generative AI In Autonomous Finance

In contrast, high-volume simple tasks do not need a specialized team of cross-experts. Or when you have a hard restriction on the number of available resources, swarming is ruled out by default. Cases that need niche expertise do not benefit from the broad expertise offered by multifunctional swarm teams. As with most traditional business relationships, sometimes the customers themselves opt out of swarming services – sticking instead with the conventional single-point-of-contact approach.

AI-Powered Swarming

As with anything today, there’s always a use case or two for adding AI to the mix – and swarming is no different. With AI, you can implement chatbots, self-service options, load balancing, case routing, task allocation, predictive analytics, knowledge base updates and more. And the outcomes are huge wins not just for the customers but also for the support reps in terms of effectiveness and speed. And in the process, the AI models are trained for better results.

ForbesSAP BrandVoice: GenAI Drives Delicious Sustainable Food Value Chain Trends

This brings us to the Intelligent Swarming methodology (developed by the Consortium for Service Innovation) which allows you to efficiently route work to the right experts and facilitate smart connections. These connections ensure skill utilization and help tap collective knowledge – for swifter, smoother resolutions. Adopters of this methodology are equipped to improve relevance, reach, and collaboration.

Use Case for Enterprise Customers

Large enterprises are characterized by complex, multilayered landscapes, and in this scenario, collaboration is key for ‘tackling the ticket’ in customer support. Swarming helps bring down ticket transfers and resolution times – two very critical requirements.

For example, even a small issue that hinders a business user from creating a dispatch could result in multiple downstream supply chain disruptions, which could have a huge impact when it comes to key industries like pharmaceuticals or utilities. Such supply chain disruptions are cross-functional in nature. In this case, a flexible model like swarming can bring the right experts together, allowing for quicker root cause analyses and resolutions – saving time and potential costs for the customer as well as the vendor.

ForbesSAP BrandVoice: 2024 Trends: Four Ways GenAI Transforms Enterprise ERP Applications

For example, as a market leader in enterprise application software, SAP uses the Intelligent Swarming methodology and the iSwarm technology to help connect experts in customer support. iSwarm uses AI to find the best experts to solve complex cases. Additionally, the integration with Microsoft Teams, allows swarm creators and responders to document their collaboration for future reference.

An AI-based Experience Engine (Expert Finder) helps bring the right people together. This engine runs through a large amount of support data to identify the best experts to be included in the swarm. This collaborative model not only builds teamwork but also allows support reps to recognize colleagues who support them.

The ‘Swarmageddon’ of Complex Cases

One might ask why swarming should matter to the customer at all. Does the customer really care how a case is resolved? The answer is ‘yes’. Businesses look for vendors and service providers that offer swift case resolutions and fewer escalations. The customer wants to see immediate and positive outcomes – and AI-powered swarming is achieving exactly that. The real question should be: who wouldn’t want an efficient closure of their most troublesome support tickets?!