Bright Data: web data collection for marketing agencies

Which marketing agencies is Bright Data a fit for?

Bright Data fits agencies that need search, competitor and AI-answer data at a scale or granularity their SEO and ad platforms do not provide — many cities, many clients, or data those platforms do not collect — and that want raw data in their own reporting rather than another dashboard.

It is not an SEO platform: there is no keyword database, rank-tracking dashboard or client portal, so the agency supplies the keyword list and the reporting layer. Teams that build client reporting in a warehouse, a BI tool or a spreadsheet gain a data source that returns the search result a user in a given city sees, on demand. The company reports 50,000+ customers on its pricing pages.

How do agencies track rankings across multiple cities with a SERP API?

A SERP API is a service that runs a search query on a search engine from a chosen location and device and returns the result page as structured data — organic results, ads, local packs, People also ask, AI Overviews — so rankings can be tracked programmatically.

Multi-city tracking is the agency case where a SERP API earns its place. Google localises results by the searcher's inferred location, so a client's rank in Lyon and in Lille can differ for the same keyword, and a single national check hides that. Bright Data's SERP API accepts a location parameter (country, state, city or postcode) and a device parameter, and returns JSON, HTML or Markdown, including the local pack — the three-business block above organic results — so local-pack positions can be tracked per postcode, the grid-style approach local SEO agencies use; Bright Data lists seven engines including Google, Bing, DuckDuckGo and Yandex, coverage in 195 countries, and pricing on successful requests only (brightdata.com, 12 September 2026). Google's documentation confirms that results vary by the user's location.

Two practical points. First, the free tier: 5,000 free credits a month, one credit per SERP request, shared with the Unlocker API, Scraper APIs and Scraper Studio, with no card required — per Bright Data's free-tier documentation, checked 12 September 2026. Second, cost at scale: pay-as-you-go is listed at $1.50 per 1,000 requests, with a $499-a-month Scale plan at a lower rate (brightdata.com, same date); 500 keywords × 10 cities × daily is roughly 150,000 requests a month.

How do you automate ranking reports for clients?

An automated ranking report is a scheduled pipeline that collects rank data, stores it with a date, computes movement against the previous period, and renders it for the client — four steps, each of which can be automated.

• Collect. A scheduled job sends each keyword × city × device combination to the SERP API and stores the position, URL and result-page features.

  This is the step agencies most often still do by hand; it is the one the API automates.

• Store with a date. One row per keyword, city, device and day, in a spreadsheet, a database or a warehouse.

  Movement reports need history; a snapshot cannot show a trend.

• Compute movement. Compare to the prior week or month; flag gains, losses and new entrants. Google's Search Console API supplies the client's own impressions and clicks to set beside external rank data.

  Rank alone does not show traffic; pairing the two is what clients ask for.

• Render. A BI tool or templated document reads the table and produces the client report on a schedule — under the agency's own branding, since the raw data carries no third-party tool's name.

  The rendering layer is the agency's; the data layer is the API.

What is a residential proxy, and how do ISP Proxies differ?

Bright Data's Residential Proxies are rotating IP addresses sourced from opt-in peer devices — real consumer devices whose owners have agreed to share connectivity through the Bright SDK — while ISP Proxies are a separate product: static IP addresses allocated by internet service providers, held for as long as the account needs them and priced per IP rather than per gigabyte. They solve different agency problems.

Rotating residential IPs suit collection that needs a different consumer IP across many locations — rank checks, ad verification, geo-restricted landing pages. Bright Data lists 400M+ monthly residential IPs in 195 countries with country, state, city and postcode targeting. Static ISP IPs suit tasks that need the same address for a whole session; Bright Data lists 1.3M+ ISP IPs, priced from $1.30 per IP (brightdata.com, 12 September 2026).

Pricing. Residential Proxies list at $8 per GB pay-as-you-go, with $499, $999 and $1,999 monthly plans that include 141, 332 and 798 GB respectively at lower per-GB rates. On 12 September 2026 the pricing page displayed a temporary 50% discount for three months, applied with coupon code RESIGB50 at checkout; treat it as a promotion that can change or end, not as list price.

Which Bright Data products require KYC?

Access to the Residential and Mobile proxy networks is subject to a KYC (know-your-customer) compliance process and use-case approval before use. Bright Data's pricing FAQ states that a representative asks each customer to complete a short compliance process that may include an introductory video call and verification of company or personal details; the Unlocker API and datacenter products do not require it (Bright Data residential pricing page, checked 12 September 2026). Plan a few working days for this before a proxy-based project starts.

How do agencies collect web data without being blocked?

Sites protect themselves with rate limits, CAPTCHAs and anti-bot systems, and the reliable way for an agency to collect public web data is to use tools built for lawful collection, keep request volumes reasonable, and stay within the provider's acceptable use policy.

Bright Data packages the handling in its own products. The Unlocker API takes a URL and returns the page, managing IP selection, retries and CAPTCHA resolution on its side; the SERP API does the same for search engines; the Browser API provides a remote browser for pages that need JavaScript, connected from existing Puppeteer, Playwright or Selenium code. All three are billed on successful responses, so a failed request is not charged (brightdata.com pricing pages, 12 September 2026). The practical effect for an agency is fewer gaps in scheduled collection and less engineering time on retries, not a guarantee that every request succeeds.

Three practices keep collection on the right side of the target site and the provider: public pages only, no content that requires authentication; request rates and daily spend limits set in the dashboard; and the use case confirmed against Bright Data's Acceptable Use Policy, which excludes categories such as account creation and collection of private data. A change of use case after KYC should be raised with the account manager.

How can an agency track whether AI answers mention a client?

AI assistants now answer a share of buying questions directly, and an agency can measure whether a client is named, cited or shown in a shopping card by running the client's buying prompts through the assistants on a schedule and recording the answers as structured data.

Google's AI Overviews can be captured with the SERP API's AI Overview parameter inside the same city-level requests used for rank tracking, so a client's presence in the AI answer can be logged alongside their organic position. For the assistants themselves, Bright Data's LLM Scrapers query ChatGPT, Perplexity, Gemini, Google AI Mode and Bing Copilot via API or without code and return the user-facing answer with parsed citations, competitor mentions and shopping cards, with country-level targeting (brightdata.com product page). In Cllimber's August 2026 test for its Bright Data ecommerce article, four US buying prompts run through the ChatGPT Search scraper returned one record of 40 fields each at 100% success. In AIMultiple's LLM scraper benchmark (1,000 tests per provider, read 28 September 2026), Bright Data's success rate was about 98% on ChatGPT and about 100% on Perplexity, Gemini and Google AI Mode; the benchmark notes Bright Data is among AIMultiple's subscribers. A managed version exists as the Bright Insights GEO add-on.

What else can marketing agencies collect with Bright Data?

How much does Bright Data cost for a marketing agency?

Verified 12 September 2026. All figures below are Bright Data's own published figures, read on brightdata.com and its documentation on the date shown. Promotions are marked; they can change or end.

What are the legal considerations for agency data collection?

Whether a given collection project is lawful depends on the data collected, the purpose, the method, the jurisdiction and third-party rights — there is no blanket rule that public-web scraping is permitted, and agency projects can involve personal data even when the target is a search result or a product page.

• Personal data. Search results, reviews and social posts can contain names, handles and other personal data. Under GDPR that data needs a lawful basis, a stated purpose and minimisation; the French regulator CNIL has published guidance on web scraping and legitimate interest. Exclude or anonymise personal fields the reporting does not need.
• Terms and database rights. These vary by site and country; reproducing facts for analysis is treated differently from copying content.
• Method. Reasonable request rates, public pages only, no circumvention of access controls, and a provider whose sourcing and acceptable-use rules are documented.

Bright Data publishes its compliance framework, certifications and peer-sourcing model in its Trust Center. That framework does not replace the agency's own legal review of a specific project. This is general information, not legal advice.

How should an agency start using Bright Data?

• Pilot on the free credits. Create an account on brightdata.com and run a keyword list through the SERP API for two or three cities.
• Start KYC early if the project needs Residential or Mobile proxies; approval is tied to the described use case.
• Set spend limits per zone in the dashboard before scheduling anything.
• Store every result with a date from day one, whatever the reporting layer.

Next step

The free monthly credits cover a rank-tracking pilot for one client across a few cities; proxy projects start with KYC. Pricing and documentation are on brightdata.com.