AI training for product managers: making sense of customer feedback

Build the program around the team's questions

Syntalith proposes team training tailored to the material product managers use. With the product lead, we discuss recurring planning questions and where the team interprets customer feedback differently. We choose exercises around the tools, conversation records, and meeting briefs already in use. The team identifies material to work with and explains the context of customer conversations.

The proposed program covers defining a research question, organizing feedback by customer and situation, and separating a quote from the author's interpretation. Further work involves comparing themes with existing product evidence and framing questions for future conversations. AI helps group material and prepare drafts. Product managers practice writing a concise rationale that lets readers return to the original feedback and see what supports the conclusion.

Feedback on a draft identifies where the reasoning needs more work. If a customer asks for a feature and the brief immediately describes a solution to build, it needs a question about the task the customer wants to complete. If “important for sales” takes the place of context from a particular conversation, the author needs to recover that context from the colleague who spoke with the customer. Review covers both the selection of information and revision of the text the team will read.

One request across several channels

Suppose the same customer asks for a data export in a support ticket, a sales meeting note, and an interview. The AI summary presents those records as evidence of widespread demand. Yet all three records concern one customer. Without that context, the team could read the frequency of recorded requests as evidence of independent needs.

The revised brief links the ticket, sales note, and interview to the same customer account. It retains the references, describes the export need, and states that the material does not show how often other customers encounter it. A useful follow-up question is: “What task does the customer need the export for, and how do they complete it today?” The importance of that need to the commercial relationship can be discussed separately from its prevalence.

Write a brief that supports the product discussion

Several roles contribute to this work. The product manager chooses the question and prepares the rationale. A researcher helps interpret the interview context, while a support or sales colleague explains the contact history. An engineering reviewer can point out where the description of a need has already become an assumption about implementation. We decide together which review sessions would benefit from their input.

The program also includes comparing different comments about a similar feature. Authors look for a shared problem while preserving differences in customers' tasks. A concise brief should explain what needs further investigation and which information is missing from the priority discussion. In an independent task, another customer describes an export for a different purpose, so the product manager updates the account of customer needs without assuming the two requests mean the same thing. The product lead can assess the work by checking whether the rationale is clear and whether they can identify a question that still needs investigation.

Use the tools already available

Jira Product Discovery documentation describes capturing research and interviews, along with links to support tickets and sales opportunities, as insights. The same insight can be copied to different ideas. These storage and linking features can also support training exercises.

If the difficulty is finding known feedback, better use of the current tool may be enough. Training focuses on interpreting the material and communicating what the team found. If a separate tool is needed to assign topics across a large collection of comments, see our article on custom models for B2B product feedback analysis.

Fit the training to the team

We consider product experience and AI familiarity separately. Someone new to the tool can work with a short conversation and a brief of their own. A confident AI user can work with more complex material from several channels, preserving links to customers and the context of their comments.

Preparation, draft review, and any work between sessions are agreed in the program. Company feedback must be approved for use in the chosen tool; exercises can use fictional material.

In an initial conversation with Syntalith, describe customer feedback that led to disagreement within the team. We will discuss whether the difficulty lies in interpreting the comments, connecting records, or writing the rationale, then use that context to shape the workshop. Further service information is available on Syntalith's pricing page.