Skip to main content

Researcher Agent

Introduction​

The Researcher Agent is your gateway to detailed, data-driven insights about any industry, market, or topic. Designed for business professionals in sales, marketing, and strategy, this tool transforms complex data into actionable reports. Below, you’ll find how to configure the Researcher Agent and unlock its full potential.

Quick Configuration Overview​

Here’s a snapshot of the main settings you can adjust. Each one can influence the style, depth, cost, and time taken by the Researcher Agent to produce your report.

  1. Creativity

    • Range: 0 – 11
    • Why It Matters: Lower values (around 0) keep the report factual and straightforward. Higher values (around 1) encourage creative, out-of-the-box ideas.

  2. Number Of Web Search Results

    • Range: Up to 20
    • Why It Matters: Determines how many search results to gather for each subtopic. Ten results are often ideal for a balanced view.

  3. Total Report Words

    • Recommended: 4000 words (or your chosen length)
    • Why It Matters: Sets an approximate target for how detailed the final report should be.

  4. Report Format

    • Choices: APA, MLA, Chicago Style, IEEE, Harvard, Vancouver, ASA, Oxford
    • Why It Matters: Each format follows different citation and presentation styles. Pick the one that best fits your needs.

  5. Researcher Instructions (Custom Instructions)

    • Example: “Explore emerging trends in [Your Industry], analyze top competitors’ strategies, and identify consumer behavior shifts in [Target Market] for [Year].”
    • Why It Matters: The clearer and more specific your instructions, the more tailored and accurate your final report will be.

  6. AI Model

    • Why It Matters: You can choose different AI engines for the research process. Some are faster and more cost-effective, while others offer deeper analysis and potentially richer insights.

  7. Maximum Response Tokens

    • Why It Matters: Tokens are like a budget for words. The higher the limit, the more detailed (and potentially longer) the final report. However, more tokens also increase the cost.

  8. Max Iterations

    • Range: 1 – 10
    • Why It Matters: Controls how many rounds of web searches the agent will conduct. More iterations provide deeper research but also use more credits.

Understanding Pricing and Credits​

The Researcher Agent calculates credits based on three main factors:

  1. Input Tokens

    • These are the tokens used to process web results.

  2. Output Tokens

    • These are the tokens in the final report. Longer, more thorough reports require more output tokens.

  3. Web Searches (tied to Max Iterations)

    • Each iteration involves web queries and data processing. The higher the Max Iterations, the more web searches and thus the higher the credit consumption.

Tips for Managing Costs:

  • Keep your Max Iterations lower if you only need a quick overview.
  • Adjust Maximum Response Tokens if you want a concise report.
  • Choose an AI Model that suits your budget and depth needs—some models are optimized for speed and cost, while others offer more advanced analysis.

Configuring Your Researcher Agent​

Step 1: Defining Creativity​

Adjust the creativity level (0 – 1) based on how innovative you want the report to be:

  • Lower Creativity (around 0): Produces more standard, reliable insights with minimal flair.
  • Higher Creativity (around 1): Sparks more imaginative ideas, which can be helpful for brainstorming or exploring unique angles.

Step 2: Number of Web Search Results​

Set how many results the tool should scrape and evaluate for each subtopic. Usually, 10 results give a balanced overview without overwhelming the final report.

Step 3: Total Words​

Aim for around 4000 words for a detailed yet concise report. Feel free to adjust if you need more or fewer details.

Step 4: Selecting the Report Format​

Different formats (APA, MLA, Harvard, etc.) cater to different presentation needs. Pick the one your audience expects, such as APA for academic reports or Harvard for business cases.

Step 5: Researcher Instructions​

Include any specific points you want the report to cover, such as competitor strategies, emerging technologies, or consumer behavior trends.

Step 6: Choosing Your AI Model​

Decide which AI engine is most appropriate:

  • Faster/Basic Models: Great for quicker, more cost-effective summaries.
  • Advanced Models: Provide deeper insights but may be pricier in tokens.

Step 7: Setting Maximum Response Tokens​

Determine how long or concise your final report should be:

  • Lower Limit: Short, focused insights.
  • Higher Limit: Comprehensive analysis with in-depth coverage.

Step 8: Defining Max Iterations​

Decide how many times the agent should search the web for subtopics:

  • Fewer Iterations: Quick results, lower cost, less detail.
  • More Iterations: Thorough exploration, higher cost, richer data.

Implementing Your Setup​

  1. Input Parameter Name

    • Typically named research_topic. This is what tells the AI Assistant what you want to study.

  2. Input Parameter Description

    • Provide a sentence or two describing your research goal. For example, “A deep dive into the top three digital marketing trends for 2025, focusing on consumer behavior and competitor strategies.”

  3. Return Response Directly In Chat

    • If checked, you’ll see the research report instantly in the chat.
    • If unchecked, the report is returned to the AI Assistant, which can then analyze it further or provide a summary.

  4. Review Pricing

    • Before finalizing, remember that your choice of AI Model, Max Iterations, and Maximum Response Tokens directly impacts the credits used.

Example Use Case​

Sarah’s Deep-Dive Marketing Report

  • Research Topic: “Identify key drivers of eco-conscious consumer behavior in the European market and analyze successful marketing strategies in this segment for 2024.”
  • Creativity: 0.5 for balanced yet innovative insights.
  • Number of Web Searches: 10 to gather recent studies and industry reports.
  • Total Report Words: 4000 for a comprehensive but manageable read.
  • Report Format: Harvard to align with her company’s style.
  • AI Model: An advanced model for thorough, nuanced analysis.
  • Maximum Response Tokens: 8000 to allow for a detailed breakdown of findings.
  • Max Iterations: 3 for a moderate level of depth without excessive costs.

Sarah’s final report details current trends, competitor approaches, and consumer expectations—enough to guide her eco-conscious marketing campaign.

Maximizing Your Researcher Agent Tool​

  1. Be Specific with Instructions

    • Clear, detailed prompts result in more targeted and accurate research outcomes.

  2. Use Creativity Wisely

    • High creativity is great for new ideas, while lower creativity is best for straightforward factual reporting.

  3. Balance Cost and Depth

    • Limiting Max Iterations and Maximum Response Tokens can manage costs but may reduce thoroughness.
    • Pick a suitable AI Model that matches your budget and depth needs.

  4. Format Matters

    • Make sure to choose the correct report format (APA, Harvard, etc.) so the output aligns with your audience’s expectations.

Report Formats​

Below are some popular report formats you can choose:

APA (American Psychological Association)​

  • Use Case: Social sciences, education, and business.
  • Features: Emphasizes publication date. In-text citations use author’s name and publication year.
  • Why Choose APA: Perfect for research-focused presentations where recent sources are crucial.

MLA (Modern Language Association)​

  • Use Case: Humanities, literature, and language studies.
  • Features: Emphasizes the author’s name and page number in citations.
  • Why Choose MLA: Ideal for research that involves textual analysis and citation of multiple literary sources.

Chicago Style​

  • Use Case: Broadly used in history and the arts.
  • Features: Offers Notes and Bibliography or Author-Date systems.
  • Why Choose Chicago: Highly versatile, allowing footnotes or in-text citations based on your preference.

IEEE (Institute of Electrical and Electronics Engineers)​

  • Use Case: Engineering, computer science, and IT.
  • Features: Numbered references in the order they appear.
  • Why Choose IEEE: Clear and efficient for technical topics with numerous references.

Harvard​

  • Use Case: Common in business, education, and sciences.
  • Features: Author-Date citation style, similar to APA with minor formatting differences.
  • Why Choose Harvard: Simple and flexible, making it widely acceptable across different fields.

Vancouver​

  • Use Case: Medical and scientific research.
  • Features: Numbers references sequentially in the text.
  • Why Choose Vancouver: Ideal for high-volume citations and streamlined referencing in scientific papers.

ASA (American Sociological Association)​

  • Use Case: Sociology and related social sciences.
  • Features: Tailored from Chicago’s Author-Date style for sociological terminology.
  • Why Choose ASA: Best for emphasizing social structures, theories, and data relevant to sociology.

Oxford​

  • Use Case: Often used in law and history.
  • Features: Footnotes for citations with a bibliography at the end.
  • Why Choose Oxford: Excellent for detailed commentary within footnotes, suitable for extensive legal or historical analysis.

Wrapping Up​

The Researcher Agent simplifies complex market, industry, or academic research by offering customizable settings for creativity, web searches, report length, formatting, and more. With new controls like AI Model, Maximum Response Tokens, and Max Iterations, you can tailor the balance between cost, depth, and speed according to your needs. By refining your instructions and understanding how pricing is calculated, you’ll produce thorough, actionable, and cost-effective research reports every time.