Using AI Data Stream

Prompt Presets

Use built-in AI personas to get specialized analysis perspectives on your data.

Prompt presets configure the AI’s analysis style before you ask your question. They’re like giving the AI a specific role or expertise.

What Are Prompt Presets?

When you select a preset, you’re telling the AI to approach your data from a specific perspective. A Data Analyst focuses on facts and figures, while an SEO Specialist focuses on search performance.

Each preset includes instructions that guide how the AI:

  • Interprets your questions
  • Prioritizes certain metrics
  • Formats its responses
  • Checks annotations for context

Using Presets

  1. Start a new conversation or open an existing one
  2. Click the Preset dropdown above the chat input
  3. Select your preferred preset
  4. Ask your question

The preset applies to all messages in that conversation. You can change presets mid-conversation if needed.


Built-in Presets

Data Analyst

Focus: Objective, fact-based analysis

The Data Analyst preset delivers straightforward analysis based solely on what the data shows. It avoids speculation and presents metrics precisely.

Best for:

  • Getting factual overviews
  • Presenting data to stakeholders
  • Unbiased trend analysis
  • When you want numbers without spin

Characteristics:

  • Reports only what data shows
  • Presents precise metrics
  • No speculation beyond the data
  • Always checks annotations for context on time-based metrics
  • Clean, structured output

Example output style:

Sessions increased 15.3% (42,150 → 48,600) comparing March to February. Organic search drove 62% of this growth. Top performing pages were /products/widget (↑23%) and /blog/tutorial (↑18%).


SEO Specialist

Focus: Organic search performance and rankings

The SEO Specialist preset focuses on search visibility, keywords, rankings, and technical SEO factors.

Best for:

  • Search performance reviews
  • Keyword analysis
  • Ranking opportunities
  • Technical SEO checks
  • Content optimization insights

Characteristics:

  • Prioritizes Search Console data
  • Focuses on clicks, impressions, CTR, position
  • Identifies ranking opportunities
  • Notes algorithm update impacts (checks annotations)
  • Links traffic changes to search changes

Example questions:

  • “What keywords are we ranking on page 2?”
  • “Which pages have high impressions but low CTR?”
  • “How did the latest core update affect our rankings?”
  • “What are our best opportunities to improve?”

Marketing Manager

Focus: Campaigns, conversions, and ROI

The Marketing Manager preset focuses on marketing performance, campaign effectiveness, and return on investment.

Best for:

  • Campaign performance reviews
  • Conversion analysis
  • Channel attribution
  • Budget allocation decisions
  • Marketing reporting

Characteristics:

  • Prioritizes conversion metrics
  • Analyzes traffic sources by value
  • Calculates ROI where possible
  • References campaign annotations
  • Focuses on actionable marketing insights

Example questions:

  • “Which campaigns drove the most conversions?”
  • “What’s our best traffic source for revenue?”
  • “How did the email campaign perform?”
  • “Where should we increase spend?”

Content Strategist

Focus: Content performance and engagement

The Content Strategist preset analyzes how content performs, what engages users, and what content opportunities exist.

Best for:

  • Content performance audits
  • Blog analysis
  • Engagement optimization
  • Content calendar planning
  • Topic performance

Characteristics:

  • Focuses on page-level metrics
  • Prioritizes engagement (time on page, scroll depth)
  • Identifies top-performing themes
  • Finds content gaps
  • Suggests content opportunities

Example questions:

  • “Which blog posts get the most engagement?”
  • “What topics drive the most traffic?”
  • “Which pages have high traffic but low engagement?”
  • “What should we write about next?”

Executive Summary

Focus: High-level KPIs for stakeholders

The Executive Summary preset delivers concise, business-focused insights suitable for leadership.

Best for:

  • Board reports
  • Weekly/monthly summaries
  • Stakeholder updates
  • Quick status checks
  • Non-technical audiences

Characteristics:

  • Simple, clear language
  • Key metrics only
  • Significant changes highlighted
  • Business impact focus
  • Actionable summaries
  • No technical jargon

Example output style:

March Performance Summary

Overall: Strong month with 12% revenue growth.

Key wins:

  • Revenue up $45K vs February
  • Mobile conversion rate improved 8%
  • Email campaign delivered 3.2x ROI

Watch items:

  • Bounce rate trending up on mobile
  • Organic traffic flat (algorithm update impact)

Ecommerce Analyst

Focus: Sales, revenue, and purchase behavior

The Ecommerce Analyst preset specializes in online store metrics and sales performance.

Best for:

  • Revenue analysis
  • Conversion funnel review
  • Product performance
  • Cart abandonment
  • Purchase behavior

Characteristics:

  • Focuses on revenue and transactions
  • Analyzes conversion funnel (view → cart → checkout → purchase)
  • Tracks average order value
  • Reviews product performance
  • Examines cart abandonment
  • Calculates ROAS when ad data available

Example questions:

  • “What was our revenue this month vs last month?”
  • “Where are people dropping out of checkout?”
  • “Which products have the best conversion rate?”
  • “What’s our average order value by device?”
  • “How much did the sale increase revenue?”

Choosing the Right Preset

Your GoalRecommended Preset
Unbiased data overviewData Analyst
Search/SEO focusSEO Specialist
Campaign performanceMarketing Manager
Content insightsContent Strategist
Stakeholder reportExecutive Summary
Sales analysisEcommerce Analyst

Combining with Questions

The preset sets the AI’s perspective, but your question drives the analysis. The same question with different presets yields different focuses:

Question: “How did last month perform?”

PresetFocus of Response
Data AnalystAll metrics, objective comparison
SEO SpecialistSearch rankings, keyword changes
Marketing ManagerCampaign ROI, conversion sources
Executive SummaryKey wins and concerns only

Creating Custom Presets

You can create custom presets tailored to your needs.

Creating a Preset

  1. Go to Settings → Presets
  2. Click Create Preset
  3. Enter a name and description
  4. Write your system prompt (instructions for the AI)
  5. Save

Writing System Prompts

Your system prompt tells the AI how to behave. Include:

  • Role: What persona should the AI adopt?
  • Focus: What metrics/areas to prioritize?
  • Style: How should responses be formatted?
  • Rules: Any specific behaviors or restrictions?

Example custom preset:

You are a technical SEO specialist analyzing website performance data.

Focus on:
- Core Web Vitals and page speed
- Crawlability and indexation issues
- Mobile vs desktop performance gaps
- Technical barriers to ranking

Always check for:
- Pages with slow load times
- Mobile usability issues
- 404 errors or broken links

Format responses with clear sections and actionable recommendations.

Sharing Presets

Custom presets are available to all team members with access to the property.


Presets and Annotations

All presets instruct the AI to check annotations when analyzing time-based data. This is why keeping annotations up to date matters.

When you ask “Why did traffic drop last week?”, the AI:

  1. Pulls the traffic data
  2. Checks annotations for that period
  3. Incorporates any notes into the analysis

Without annotations, the AI can only speculate. With annotations, it provides context.


Tips for Using Presets

Match Preset to Audience

  • Presenting to the CEO? Use Executive Summary.
  • Working with the SEO team? Use SEO Specialist.
  • Analyzing campaigns? Use Marketing Manager.

Switch Presets for Different Views

In one conversation, you can:

  1. Start with Data Analyst for the facts
  2. Switch to SEO Specialist to dive into search
  3. End with Executive Summary for the stakeholder version

Don’t Over-Rely on Presets

Presets are helpful, but a well-asked question matters more. A specific question to a generic Data Analyst often beats a vague question to a specialized preset.

Next Steps