BizTrend AI Analyst

What does BizTrend AI Analyst do?

BizTrend AI Analyst turns natural-language questions into actionable Amazon product research. Ask about products, keywords, organic or sponsored positions, prices, ratings, review counts, Best Sellers Rank, historical changes, comparisons, and BizTrend Scores.

The Actor returns a concise answer together with the products and dates that support it. Run it manually in Apify Console or connect it to your workflows through the Apify API, schedules, webhooks, and integrations.

Why use BizTrend AI Analyst?

• Research product opportunities without writing code or queries.
• Find products by keyword, attributes, or a real-world use case.
• Compare multiple ASINs using consistent product data.
• Review keyword positions and ranking changes over time.
• Explore product history and Best Sellers Rank history.
• Discover high-scoring or fast-moving products.
• Receive clear warnings when data is missing or incomplete.
• Use your own OpenAI API key to control AI usage and billing.

The Actor answers only from available product data. It reports gaps instead of inventing unavailable prices, rankings, scores, or history.

How to use BizTrend AI Analyst

1. Open the Actor's Input tab.
2. Enter one question in Question.
3. Paste your OpenAI API key into the protected OpenAI API key field.
4. Optionally select a marketplace and adjust the maximum number of results.
5. Click Start and open the Output tab when the run finishes.

Example questions:

• Find 20 organic products for the keyword wireless mouse on amazon.co.uk
• Compare B0EXAMPLE1 and B0EXAMPLE2 on amazon.com
• Find products under $30 with at least a 4.5 rating on amazon.com
• Which products are useful for organizing cables while traveling?
• Show the 30-day Best Sellers Rank history for B0EXAMPLE1
• Which products currently have the strongest momentum?

When a supported Amazon hostname such as amazon.co.uk appears in the question, the Actor uses it automatically. A marketplace selected in the Input tab takes priority. If neither is provided, the default marketplace is amazon.com.

Input

The simplest input contains a question and your OpenAI API key:

{
    "query": "Find products for a minimalist desk on amazon.co.uk",
    "openAIApiKey": "YOUR_OPENAI_API_KEY"
}

Field	Type	Required	Description
query	string	Yes	One natural-language Amazon product research question.
openAIApiKey	string	Yes	Your OpenAI API key. It is handled as a protected secret input.
region	string	No	Amazon marketplace override. It takes priority over a hostname in the question.
limit	integer	No	Maximum returned products, from 1 to 50. Default: 10.
includeEvidence	boolean	No	Include source labels supporting the answer. Default: true.
includeNearMatches	boolean	No	Allow explicitly labelled near matches without relaxing structured filters. Default: false.
forceRefreshKeyword	boolean	No	Request one fresh collection for an explicit keyword question. Default: false.
autoEnrich	object	No	Controls bounded collection of missing or insufficient product data.

The number written in the question does not override limit. For example, a request for 100 products still returns no more than the configured maximum of 50.

Advanced data collection

The default autoEnrich settings are suitable for most users:

{
    "missingAsins": true,
    "missingKeywords": true,
    "refreshStaleKeywords": false,
    "keywordFreshnessHours": 168,
    "noResultRetryHours": 168,
    "maxRuns": 1,
    "maxKeywordsPerRun": 1,
    "maxPages": 10,
    "maxProducts": 50
}

Reduce these limits when you want a faster, lower-cost run. Set maxRuns to 0 to use only currently available data. Forced refresh and larger collection budgets can increase run duration and cost.

Output

Each run writes one analysis to the default Dataset and the OUTPUT record:

{
    "status": "success",
    "query": "Find products for a minimalist desk on amazon.co.uk",
    "answer": "Here are the most relevant products available for this request...",
    "products": [
        {
            "asin": "B0EXAMPLE1",
            "title": "Example desk accessory",
            "brand": "Example Brand",
            "price": 24.99,
            "currency": "GBP",
            "rating": 4.6,
            "reviewCount": "1250",
            "productUrl": "https://www.amazon.co.uk/dp/B0EXAMPLE1"
        }
    ],
    "evidence": {
        "dataAsOf": "2026-08-14T08:00:00.000Z",
        "historyCoverageDays": null
    },
    "dataGaps": [],
    "clarification": null,
    "errors": []
}

You can download the Dataset in formats such as JSON, CSV, Excel, XML, RSS, or HTML, or consume it directly through the Apify API.

Data table

Field	Description
status	Completeness of the result, such as success, partial, no_results, or insufficient_data.
query	The original question.
answer	A concise, evidence-based response.
products	Products and attributes supporting the answer. Available fields vary by question.
evidence.dataAsOf	Most recent timestamp among the data used.
evidence.historyCoverageDays	Available historical coverage when the question requires history.
dataGaps	Missing, limited, or unavailable information that affects the answer.
clarification	A follow-up question when more detail is required.
errors	Sanitized run errors, without credentials or raw stack traces.

Common status values:

• success — the available data supports the answer.
• partial — useful results exist, but some requested data is incomplete.
• no_results — no matching products or observations were found.
• insufficient_data — data exists, but there is not enough history or coverage for the requested analysis.
• clarification_required — the question needs an ASIN, keyword, marketplace, or other detail.
• unsupported — the request falls outside the Actor's product-analysis capabilities.
• error — the run could not complete because a required service or configuration was unavailable.

How much does it cost to analyze Amazon products?

Cost depends on Apify compute usage, OpenAI usage billed to your API key, the requested result limit, and whether additional product data must be collected. Your available credits and free usage depend on your current Apify and OpenAI plans.

For lower cost and faster runs:

• Keep the result limit close to the number of products you actually need.
• Ask one focused question per run.
• Use an ASIN for product-specific questions.
• Avoid requesting long historical periods unless they are necessary.

Tips and advanced options

• Use the full marketplace hostname, for example amazon.co.uk, when you do not select a marketplace override.
• Include the word keyword and specify organic or sponsored for search-result questions.
• Use valid 10-character ASINs for product details, comparisons, and history.
• State numeric filters explicitly, such as under $25, rating at least 4.5, or more than 500 reviews.
• Describe the intended use when searching conceptually, such as products for organizing charging cables in a backpack.
• Smaller ranking numbers are better. Sponsored results do not have an organic position.
• Recently discovered products may not yet have complete scores or historical coverage.
• Strict filters are not silently relaxed. Fewer products may be returned when only a limited number satisfy the request.

FAQ, privacy, disclaimers, and support

Is my OpenAI API key included in the output?

No. The key is submitted through a protected secret input and is not included in the Actor's Dataset, OUTPUT record, or application logs. Do not place API keys or other credentials inside the question itself.

Why did I receive fewer products than requested?

The Actor respects the limit setting and only returns products supported by the available data and requested filters. Amazon may provide fewer matching results, and the Actor does not add unrelated products just to fill the requested count.

Why is product history or a score missing?

Some products do not yet have enough observations for historical analysis or a complete score. In those cases, the Actor returns partial or insufficient_data and explains the limitation in dataGaps.

Can the Actor answer general questions unrelated to Amazon products?

No. It is designed for Amazon product research, keyword positions, rankings, history, comparisons, and BizTrend analytics. Unrelated questions return an unsupported status.

Is scraping Amazon legal?

You are responsible for complying with Amazon's terms, robots policies, applicable laws, and data-use restrictions. Do not use this Actor to collect personal data, protected content, or prohibited information.

For bugs, feature requests, or questions, use the Issues tab on the Actor page. Custom workflows and output formats may be available as a custom solution.