> ## Documentation Index
> Fetch the complete documentation index at: https://knowledge.deeto.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Insights: Best Practices

> How to write better queries, apply AI Insights across your team, create content assets with Content Studio, and build habits that make customer intelligence a daily practice.

<Info>
  This guide covers how to get the most out of AI Insights — also called the Personal AI Assistant in the platform. They are the same feature. Not sure what AI Insights is yet? Start with the [AI Insights overview](/docs/platform-guides/ai-insights).
</Info>

AI Insights gives your team access to all your Voice of Customer data through natural language — no filters, no reports, no manual searching. This guide covers how to use it well, build team habits, and measure impact.

**Adoption journey:** Getting Started (Enabled) → Building Habits (Adopted) → Going Deep (Embedded)

***

## 1. Your first week

**Setup checklist**

* Confirm AI Insights is visible in your Deeto sidebar (navigate to **/ai-insights**)
* Verify at least 3 team members have access
* Confirm your VoC data sources are connected — interviews, surveys, reviews, and other feedback
* Run your first test query (see examples below)
* Share access with a colleague and run a query together

**Your first 5 queries**

Start with these to see immediate value:

1. "What are the top 3 reasons customers chose us in the last 6 months?"
2. "Which customers mentioned challenges with onboarding?"
3. "Find positive quotes about \[specific feature] from enterprise customers"
4. "What feature requests came up most in recent interviews?"
5. "Summarize customer sentiment about our pricing"

<Tip>
  Be specific. "What did customers say about pricing in Q4?" returns better results than "Tell me about pricing."
</Tip>

***

## 2. Writing better queries

The AI understands natural language, but a few patterns consistently produce stronger results.

**Query formulas that work**

| Pattern          | Example                                                                        | Why it works                                  |
| ---------------- | ------------------------------------------------------------------------------ | --------------------------------------------- |
| Time-bound       | "What feedback did we get about integrations in the last 90 days?"             | Narrows results to recent, relevant data      |
| Segment-specific | "What do enterprise customers say about our support experience?"               | Filters by customer type for focused insights |
| Comparative      | "How does customer sentiment about Feature X compare to Feature Y?"            | Surfaces relative strengths and gaps          |
| Trend-seeking    | "Has customer satisfaction with onboarding improved over the last 2 quarters?" | Identifies patterns over time                 |
| Action-oriented  | "Find customer quotes I can use in a pitch deck for healthcare prospects"      | Produces ready-to-use output                  |

**Refining results**

If the first response doesn't quite hit the mark:

* Add a time filter: "...but only from the last quarter"
* Narrow the segment: "...specifically from mid-market SaaS companies"
* Change the format: "Can you list these as bullet points with customer names?"
* Go deeper: "Tell me more about the third point you mentioned"

***

## 3. Use cases by team

**Customer Success**

* Pre-QBR prep: "Summarize all feedback from \[Account Name] in the last 6 months"
* Risk detection: "Which accounts expressed frustration about support response times?"
* Expansion signals: "Which customers mentioned wanting to use Deeto for more teams?"

**Product Marketing**

* Competitive intel: "What do customers say about us vs. \[Competitor Name]?"
* Proof points: "Find customer quotes about ROI or time savings"
* Messaging validation: "Do customers describe our platform as 'easy to use'? What words do they actually use?"

**Product Management**

* Feature prioritization: "What are the top 5 feature requests from enterprise customers?"
* Sentiment tracking: "How has sentiment about our API changed over the last year?"
* Pain points: "What are the most common complaints about our reporting capabilities?"

**Sales Enablement**

* Deal prep: "Find customer stories from healthcare companies that implemented in under 30 days"
* Objection handling: "What do successful customers say when asked about the initial learning curve?"
* Social proof: "Find 3 strong customer quotes about our competitive advantages"

***

## 4. Creating content assets with Content Studio

Content Studio is included with AI Insights. It lets you generate polished, customer-sourced content directly from your VoC data.

| Asset type        | Best for                           | How to use                               |
| ----------------- | ---------------------------------- | ---------------------------------------- |
| Customer quotes   | Sales decks, case studies          | Pull and format verified customer quotes |
| Insight summaries | Executive briefings, QBRs          | Summarize themes across multiple sources |
| Proof points      | Marketing campaigns, landing pages | Extract specific ROI and outcome data    |
| Trend reports     | Product planning, board updates    | Identify patterns over time              |

**Best practices for content creation**

* Always verify generated content against the source citations provided
* Use specific prompts: "Create a customer quote card about our onboarding experience from a fintech customer"
* Request multiple options: "Give me 3 different proof points about our time-to-value"
* Iterate: "Make this quote more concise" or "Can you find a stronger example?"

<Note>
  Content Studio uses **40 credits per asset generated**. Plan your content needs and batch requests where possible. AI queries use **2 credits each**.
</Note>

***

## 5. Building team habits

**Weeks 1–2: Foundation**

* Each team member runs at least 3 queries
* Share one useful insight with the broader team
* Identify one recurring task AI Insights can replace (e.g. manual QBR prep)

**Weeks 3–4: Integration**

* Add "AI Insights query" as a step in at least one existing workflow
* Run a team session: each person shares their best query and result
* Identify 2–3 "go-to queries" the team will use regularly

**Month 2+: Embedding**

* AI Insights is the first stop for any customer intelligence question
* Team members proactively share AI-surfaced insights without being prompted
* New team members are onboarded to AI Insights as part of tool training
* Quarterly review: measure time savings and quality of insights

<Tip>
  The fastest way to build adoption is a quick win. Find one query that saves a team member real time in their first week, then share that story.
</Tip>

***

## 6. Measuring success

| Metric                 | Target                          | How to track                |
| ---------------------- | ------------------------------- | --------------------------- |
| Weekly active users    | 3+ users querying weekly        | Deeto usage analytics       |
| Query volume           | 5+ queries/week across the team | Platform analytics          |
| Time saved             | 5+ hours/week per team          | Before/after comparison     |
| Content assets created | 4+ per month via Content Studio | Content Studio activity log |
| Cross-team usage       | Users from 2+ departments       | User department mapping     |

<Tip>
  Share these metrics with your Deeto CSM during check-ins. They can help you benchmark against similar customers.
</Tip>

***

## 7. Troubleshooting

<AccordionGroup>
  <Accordion title="Results seem incomplete">
    Check that all your data sources are connected. AI Insights can only search data that's in your Deeto workspace. If you've recently imported new content, allow a short processing delay before querying it.
  </Accordion>

  <Accordion title="Responses are too generic">
    Add more specificity — time range, customer segment, specific topic. See the query formulas in Section 2. "What do customers say about onboarding?" is weaker than "What do enterprise customers say about the first 30 days of onboarding in the last 2 quarters?"
  </Accordion>

  <Accordion title="Team members aren't using it">
    Start with a demo focused on their specific pain point. One relevant query is worth more than a feature overview. Find the task they do manually every week and show them how AI Insights does it in 30 seconds.
  </Accordion>

  <Accordion title="Running low on credits">
    AI queries use 2 credits each. Content Studio uses 40 credits per asset. Talk to your Customer Success Manager about credit allocation and planning.
  </Accordion>

  <Accordion title="Want to expand to more users">
    Contact your Customer Success Manager to discuss additional seats and access configuration.
  </Accordion>
</AccordionGroup>

***

## Support

Contact your Deeto Customer Success Manager or email [support@deeto.ai](mailto:support@deeto.ai).

<CardGroup cols={2}>
  <Card title="AI Insights" icon="brain-circuit" href="/docs/platform-guides/ai-insights">
    What AI Insights is, how it works, capabilities, limitations, and FAQs.
  </Card>

  <Card title="Research Hub: Best Practices" icon="flask" href="/docs/platform-guides/research-hub-best-practices">
    For structured research projects, cadence, and stakeholder reporting.
  </Card>
</CardGroup>