Insights
Surface clusters of insights around requests, defects, praises, and learnings
Insights
The magic of Inari starts when generating new customer and product insights. Think of Insights
as the top trends surfaced from all of your customer interactions, specifically around top Requests
, Defects
,Praises
, and Learnings
.
In the nav bar, you can view Insights
that are tracked under each Space. After uploading feedback for the first time, Inari will automatically analyze it, extract and categorize all the top customer quotes, then cluster them and present them as new insights here.
Every insight has more details that can be viewed by clicking on the Insight
column. This shows the detail panel with a deep-dive, all Highlights
that were automatically linked to it based on keyword and semantic similarity, volume and sentiment metrics in Trends
, and linked Issues
tracked in the backlog.
Insight Statuses
Every Insight
in Inari has 2 potential statuses to make them easier to manage:
- Trending: trending insights are generated on-the-fly as new feedback gets added into Inari or are created based on a user’s explore queries. Trending insights disappear if they aren’t saved as new feedback and queries are run.
- Saved: insights that teams can manually save if they want to monitor the insight moving forward and keep metrics updated when new feedback gets added.
Explore Trending Insights
When teams add new feedback into Inari, trending insights show the latest emerging trends based on clusters identified in that recent feedback. If you think a trending insight is useful and want to monitor it, you can save it by pressing Save Insight.
Trending insights disappear whenever new “explore” queries are run or a new set of feedback is added into Inari.
If you press Explore Insights, you can set filtering and prompting criteria to generate new insights based on your requirements. Most importantly, you can filter for specific date ranges, keywords, types of feedback, and other fields for limiting the set of feedback you’re analyzing. You can also provide a custom prompt or adjust the volume of insights you want to generate to set the insight’s granularity level.
Trending insights are optimized for quick, iterative searches to make it easy to identify a wide of potential topics then save the set of insights you want to track moving forward.
Generate Saved Insights
You can generate Saved Insights
by saving insights from your Explore Insight
queries OR you can manually trigger the clustering by clicking Automations
then Generate New Insights
- this will look for Feedback Highlights
that haven’t previously been analyzed, cluster them together, then surface the top trends.
Once saved, Inari will generate a more accurate and robust search across your connected feedback, generate a deep dive, and provide more accurate metrics. You can edit the summaries yourself in the detail panel, see relevant attributed highlights, and review the metrics in the Trends
tab.
Citations
Every detailed description tied to Trending
and Saved
insights will have Citations
. Each sentence within the description will be linked back to feedback highlights that are semantically related to it.
The number within each citation is the volume of feedback highlights related to it. You can press the citation to filter the feedback highlights to ones that are most relevant to that sentence then press the chat icon to remove the filter afterwards.
Attributing Highlights to Insights
As new insights are created and new feedback highlights are created, Inari automatically links feedback highlights to insights based on semantic and keyword similarity. This enables each insight to have useful volume metrics, sentiment scores, and a clean repository of linked customer quotes to serve as context for backing the insight.
This linking automatically happens whenever new feedback highlights are created but can also be manually triggered by pressing Attribute Highlights
in the Insights
page.
You can also customize keywords or phrases per insight in the detail panel so that if an exact keyword match exists on a feedback highlight, it gets auto-linked to the insight.
Generate Deep Dive
Inari generates a Deep Dive
for every insight based on the topic of the insight and the context gathered from all linked customer quotes.
This deep dive is auto-generated after the insight is initially created but is refreshable by pressing the Last Updated
or Generate Deep Dive
buttons.
Generate Issues
Inari also generates backlog Issues
for insights by taking the problem identified in the insight, context gathered from all linked customer quotes, and external context about the org to generate potential actions a product team might take to resolve this issue.
You can manually create these issues by pressing Generate Issues with AI
in the Issues
tab in the detail panel.
Prioritize Insights
Each insight has customizable priority levels that are set by AI based on the context from the insight, the volume and severity of the metrics linked to the insight relative to other insights in that space, and general product heuristics on how best to prioritize problems.
Priority levels can be manually overridden or re-prioritized in bulk by pressing Automations
then Prioritize Insights
.
Insight Automations
In the Settings
page and Insights
tab, you can toggle several insight automations per space:
- Auto-search for new insights as more feedback is added and triaged into the space.
- Set which types of insights are automatically surfaced (ex: Requests, Defects, Praises) as new feedback highlights are created.
- Provide custom instructions to Inari’s system prompt on how priority levels are set.
- Creating alerts to push into Slack when new insights are created or high volumes of customers are being attributed to an existing insight.