In part one of this series, we took a look at how artificial intelligence (AI) can be used to elevate managed service provider (MSP) efficiency and improve MSP operations. Here, we are going to look at how AI can improve an MSP’s margins through customer management and sales enablement.
One of the areas where MSPs can leverage AI is to identify orphan usage. With many accounts, the customer is a bit lackadaisical in managing things to the required degree. Additions, changes, and deletions (ACDs) of individual accounts within the overall account may not happen as often as they should. AI can identify accounts that do not appear to be in use and can suspend them entirely, thereby saving resources.
The MSP can then approach the customer to see if they need help in managing ACDs. This is only a temporary means of improving margins. The fact that the MSP is then seen by the customer as an ‘honest partner’, helping them to save money by reducing unnecessary costs through over-licensing, can lead to greater customer stickiness and a chance for additional services further down the line.
AI-driven customer management and sales enablement
AI can also manage customers more effectively, going beyond mere advice on ACDs (Automatic Call Distributors). An MSP with a large number of customers can use AI against a broad set of anonymized data from all customers to identify where a specific customer may be running things sub-optimally or may be showing signs of needing access to further services to meet their desired outcomes.
Instead of tasking account managers with identifying these areas, AI can swiftly pinpoint sales opportunities and transfer them to the account manager for further action. Again, generative AI is now at a point where it is possible to create a basic guide for a customer to apprise them of where their issues may be, so pre-educating them before the account manager gets in touch. This also minimizes the expensive (and, unfortunately, error-prone) human input in the process, thus providing opportunities for margin improvement.
Streamlining customer interactions
Finally (for this blog, at least), comes customer automation. Nowadays, users primarily interact with generative AI systems through textual input. Voice input is becoming increasingly widespread outside of the consumer systems of Siri, Alexa, and so on. The majority of customers are not technically savvy enough to know how to put into words exactly what they want. For example: “I’d like a container with a virtual CPU corresponding to a Xeon 6740E, with 32GB equivalent memory, and 100GB of initial storage, with the capability to expand all to deal with expected overheads of up to 40 percent.”
Using a simple textual interface between the customer and AI, this could be simplified to more meaningful (to the customer) requests, such as “Based on our organization having 30 users and 2,000 customers, could you please set us up with a sales system to manage 500 products?”
AI can then use data from large language models (LLMs) combined with more focused task-specific data (based on what the MSP can do) to decide that the customer needs the virtual Xeon system mentioned above – again, without any human input. Using event processing, it can even kick off the provisioning of the required solutions directly. It can even lead to setting up customer billing and invoicing.
We are rapidly reaching the point where AI will become trustworthy across so many areas of an MSP’s work. The real problem then becomes whether an MSP wishes to be nice to its employees or just replace them with AI.
Photo: Shutter_M / Shutterstock
