In today's data-driven world, even the most advanced enterprises struggle to unlock the full value of their customer feedback. The flood of data from countless channels can bury critical insights, leading to misaligned product strategies and missed opportunities within competitive markets.
Productboard Pulse's conversational AI helps you cut through the noise by responding to your prompts, answering follow-up questions, and generating a report based on your filters. You can then save that report as a document to be shared across your Productboard workspace.
In this article:
Creating a report
To create an insights report using Productboard Pulse:
- Open any insights board.
- Adjust the board's filters to identify the subset of notes you want to include in the report. Feedback from any note which is excluded by the filter will not be analyzed.
- Click the Analyze button to open the Productboard Pulse sidebar.
- In the prompt field, describe how you'd like Productboard Pulse to craft the report, or choose a pre-defined prompt from the dropdown.
- (Optional) After the AI has finished writing its response, type another question into the prompt field to have it respond again. For example, you could ask it to identify the most important pain points out of the ones it listed previously.
Saving and sharing a report
When you're done prompting the AI and have everything you need, you can save the output as a document like so:
- Click Save at the top of the report (not the top of the board).
- The report will be saved as a new board in the same teamspace as the insights board you just analyzed, and will have a title similar to "Analysis of [original insights board]". For more information on working with documents, see Collaborate on documents within Productboard.
On citations
To boost traceability and trust, Productboard adds citations to AI-generated content. Citations appear beside each AI-generated bullet point or paragraph in both the Productboard Pulse analysis sidebar and any documents created from that analysis. When clicked, they provide clear references to the sources of insights, feedback, and data that inform your product decisions.
In the citation window, you can use the arrows to cycle through each cited note, or click Read note to open the note directly. You can always return to your point of origin using the back arrow at the top of the note panel.
By linking AI output to its original source, whether it's a customer interview, support ticket, or survey response, product teams can maintain transparency and build trust with stakeholders.
Citations also enable teams to validate the relevance and authenticity of information, ensuring that prioritization decisions are well-founded. This level of traceability fosters collaboration, alignment, and confidence across the organization, ultimately leading to better product outcomes.
Use case examples
There are tons of ways you could use Productboard Pulse's report generation. Here are two examples of how you could configure your filters to help you tackle specific use cases.
Quarterly Voice of Customer reports
Review last quarter's feedback to help you prepare this quarter's plan by setting your time filter to catch all notes created in the last 90 days. If you have segments, consider creating an insights board for each segment for comparison.
Post-release evaluations of qualitative feedback
Once enough time has passed after you release a feature, it can be helpful to review what people have been saying about it.
First, set your time filter to the date of the release.
Then, create a filter to help you corral notes related only to the feature in question. There are a few different ways to approach this depending on your situation
- If you're really good about linking feedback to features as it comes in, you can use a Hierarchy filter to grab all notes that have been linked to the feature you're investigating.
- If you're a little behind on linking insights but you have a robust tagging system, you can filter for tags related to the feature. This will catch notes even if they haven't been processed yet.
- If you don't have a solid tagging system down yet, you could create a Content filter and give it keywords related to the feature in question. This will surface notes that contain some, all, or none of the keywords, according to your settings.
See Filter the Insights board to find relevant feedback for more on filtering.
Current limitations
-
Our AI models work with the content of feedback notes, details about the customer who provided the piece of feedback, and segment information. Currently, the models don't yet have knowledge about other note attribute such as tags, note owner, creation date, or state.
See also
- Productboard Pulse
- Accelerate feedback analysis with Productboard Pulse
- Productboard AI vs Productboard Pulse: What’s the difference?
Comments
Article is closed for comments.