AI training for customer success: preparing a useful account review
The customer success manager has an activity summary but still does not know how to open the meeting. A list of logins and tickets does not explain what is worth discussing. AI training can give CSMs practice turning notes into a short account-review agenda: returning to the customer's goal, selecting a relevant development, and preparing a question the customer can answer.
Syntalith
Choose what to bring into the conversation
Suppose a customer wanted its branches to submit service requests through a portal. Headquarters uses it, but one branch still calls. The account notes say that branch has not yet received its planned introduction to the portal. The CSM is preparing a review using this information.
With AI, the participant drafts an agenda. The first version includes “review portal activity” and “next steps.” The instructor asks what the author wants to learn from the customer. Reviewing activity alone will not explain what is getting in the way of all branches submitting requests through the portal.
The CSM returns to the goal in the notes and prepares a more specific agenda:
- Confirm whether having every branch submit service requests through the portal is still the customer's goal.
- Discuss the branch that still calls and is waiting for its planned portal introduction.
- Agree with the customer who can help arrange that introduction and what needs to be clarified first.
Each item gives the meeting a subject to discuss. The final point does not mean the customer has already agreed to a date or named a person. Those are matters the CSM intends to work through in the conversation.
Open with something the customer can respond to
The author then uses AI to draft a few opening sentences. A sample opening for the exercise could read:
Your goal was for branches to submit service requests through the portal. Headquarters is using it, while one branch still calls in its requests. Our notes show that the branch has not yet received its planned introduction to the portal. Is that goal still current, and what do we need to work out to arrange the introduction?
A colleague plays the customer, working from the same material. They might say they understand the topic but would prefer to answer one question at a time. The instructor suggests splitting the ending: confirm the goal, pause for a response, then ask about arranging the introduction. The author revises the opening and tries the conversation again. This gives them practice using prepared text in a dialogue.
Another piece of feedback might address an earlier version that referred only to “low activity.” The instructor points out the missing connection to the customer's work: the issue is how people submit service requests, and the branch has not yet had its introduction. Restoring that context makes it possible to ask about a specific obstacle. An increase in logins alone would not establish whether branches were actually submitting requests through the portal.
During the review, ask the person playing the customer what they remember from the opening and which question they could answer. If a long account history obscured the topic, the CSM can shorten it. The agenda keeps the information that helps the two people discuss the goal and the work ahead.
Use the feedback in a later attempt
The next part of the exercise adds a new note: the branch has now received its portal introduction, but requests still come by phone. The reason has not been established. The author updates the agenda and opening using the earlier feedback. There is no need to propose the introduction again. Instead, the CSM acknowledges that it happened and prepares a question about what is making it difficult for the branch to submit requests through the portal now.
The colleague plays the customer again and checks whether they can distinguish what the CSM already knows from what they are asking. The instructor compares the two attempts: did the author use the new information, and are they giving the customer room to respond?
Use a tool's summary to prepare the discussion
If the team uses Gainsight, check the available features before settling the training program. Gainsight describes AI Cheat Sheet, a summary that draws on sources including the past six months of Timeline entries. Expanding a point provides more detail and a link to the original entry.
In our example, the CSM returns to the recorded customer goal and the plan for the branch's introduction. If an older goal is missing from the summary, it still needs to inform the conversation. This is useful practice regardless of the product: selecting a detail from the account's history that explains why a topic matters now.
Gainsight also documents Success Snapshots, presentation templates populated with customer data. If copying figures into slides takes most of the preparation time, making better use of that feature may be sufficient. A conversation-preparation workshop is relevant when the information is available but CSMs struggle to choose a topic and lead the discussion. Building a system to gather information from several sources is a separate purchase, covered in our article on an AI agent for customer contract renewal preparation.
Organize practice around the team
Someone new to AI can prepare an agenda and opening from a short account note, then revise them after a conversation in pairs. Assess tool familiarity separately from customer success experience: a seasoned CSM may also be new to AI. Someone already comfortable drafting with it can work from a longer history and practice choosing one topic for the start of the meeting. The CS leader helps establish what the company expects from an account review. An onboarding colleague can join the discussion of recorded commitments if their team arranges the product introduction.
Syntalith can tailor team training to the material and conversations your CSMs work with. A proposed workshop includes preparing an agenda with AI and practicing the opening with a colleague playing the customer. Time for revisions, another attempt, and participation by other roles are agreed in the program around the team's meeting schedule. Fictional account histories can support the exercises; real notes need to be approved for use in the chosen tool.
We also build applications and automations, so difficulty collecting the material can be examined separately. To begin, describe an account review that took a long time to prepare but still became a presentation of data. Explain what the CSM had wanted to learn from the customer. Service information is available on Syntalith's pricing page.
Plan AI training around your team’s work
Tell us about the participants, their tools and experience. We will discuss a suitable training scope and how it would be priced.
Explore team training