DataForSEO
SEOPull SEO, SERP, and keyword research data into AI workflows. Agents run organic and maps searches across Google and Bing, score keyword difficulty in bulk, expand seed terms, and surface ranking keywords for any domain.
Weekly content planning cycle starts on schedule for the marketing team
Agent reads the seed-term list from a managed source such as a Google Sheet
Agent calls Get Keyword Suggestions and Get Related Keywords to expand each seed
Agent calls Get Bulk Keyword Difficulty to score the combined keyword pool
Agent filters the pool by difficulty band and minimum monthly volume
Agent writes ranked opportunities back to the planning sheet with metrics attached
Content lead pinged in Slack with a summary of the new opportunities
Editorial review and topic approval required before any opportunity becomes a brief
What This Integration Enables
DataForSEO integration brings live SERP and keyword data into agent workflows. Agents fetch organic and Maps SERP results across Google and Bing for ranking checks and competitive snapshots. They expand seed terms with related keywords and suggestions, score difficulty in bulk so a large pool can be ranked in one call, and pull the keywords that any domain is currently ranking for. The connector now reaches well past keyword work: backlink profiles, full site audits, brand citations, technology and WHOIS prospecting, business listings and reviews, Amazon and app store listings, and a set of AI search visibility reads that answer what an assistant says about you when nobody is watching. Some of that arrives synchronously and some queues as a task you collect later, so a flow that audits a site is not the same flow that reads the result.
Without FlowRunner
One-off keyword research Researchers run keyword tools manually for each brief, with no shared dataset across the team
Stale rank tracking Competitor and target keyword positions get checked when someone remembers, not on a cadence
Difficulty scored by hand Editorial picks topics based on intuition or one-at-a-time difficulty checks
With FlowRunner
Scheduled keyword refresh A weekly run expands seed terms and updates the opportunity pool with no manual research time
Live SERP snapshots on demand Any flow can ask for current SERP results for a query the moment the question arises
Bulk difficulty grounded in data Editorial reviews difficulty-scored opportunities ranked against current search volume
Use Case Scenarios
Weekly Keyword Refresh and Opportunity Ranking
Every Monday morning, the SEO agent runs a scheduled refresh. It reads the seed-term list from a managed Google Sheet, expands each term with Get Keyword Suggestions and Get Related Keywords, and combines the result into a deduped pool. It scores the pool in one call using Get Bulk Keyword Difficulty, filters by difficulty band and monthly volume, and writes the ranked opportunities back to the sheet with all metrics attached. The content lead opens the sheet on Monday to a freshly ranked list rather than starting from scratch.
Competitor Gap Snapshot on Demand
A Slack On Mention trigger fires when a teammate posts a competitor domain in the channel. The agent calls Get Keywords For Site for that domain, picks the top-50 ranking keywords by volume, attaches the difficulty score for each, and replies in-thread with a clean table. The team gets a competitive snapshot without anyone opening a separate research tool.
Brief Generation with Editorial Approval
A new brief request lands in the content-briefs Notion database. The agent calls Google Organic Search for the working title and Get Keyword Overview for the primary keyword, then attaches the top SERP results and the keyword metrics to the Notion page. Before any of the suggested supporting keywords are added to the brief as targets, the agent posts a summary to Slack and asks the editor to confirm or revise the keyword set. Editorial owns the final call; the agent prepares the evidence.
Human-in-Loop Highlight
When an agent has scored a fresh batch of keyword opportunities and is about to add them as content targets, it does not write them into the brief silently. It posts a Slack summary to the content lead: "Refresh complete. [N] new opportunities scored. Top 10 attached. Approve adding to the content calendar or revise the filter?" The editor reviews the list and approves or trims it. The agent does the research and ranking; the editor sets the editorial direction.
Agent processes routinely
Detects exception requiring judgment
Clear match Continues automatically
Ambiguous Routes to human via Slack
Human decides
Agent resumes with decision
Agent Capabilities
55 actionsSERP
3- Google Organic Search Fetches the current Google organic SERP for a query in a chosen location and language. Used in rank tracking, competitive snapshots, and brief research where the agent needs to see what is actually ranking right now.
- Google Maps Search Fetches the Google Maps local SERP for a query, returning listings with names, addresses, ratings, and review counts. Used in local SEO and multi-location rank tracking.
- Bing Organic Search Fetches the Bing organic SERP for a query. Used by teams who track cross-engine visibility rather than assuming Google is the whole market.
Keywords Data
2- Get Keyword Search Volume Returns Google Ads monthly search volume, CPC, competition, and historical trend for up to 1,000 keywords in one call. The canonical demand signal, straight from the ad platform.
- Get AI Keyword Search Volume Returns how often keywords are used as prompts inside AI assistants rather than typed into a search box, for up to 1,000 keywords with month-by-month history. This is a different demand curve from search volume, and the gap between the two is the interesting part.
Keyword Research
5- Get Keyword Overview Returns volume, CPC, competition, difficulty, intent classification, and SERP feature data for up to 700 keywords in one call. The quick read on whether a target is worth pursuing.
- Get Bulk Keyword Difficulty Scores difficulty on a 0 to 100 logarithmic scale for up to 1,000 keywords in one call. Used to rank an expanded seed pool by opportunity instead of one term at a time.
- Get Related Keywords Expands a seed keyword into a tree of related terms using SERP-overlap analysis, with metrics on each. Depth controls how far the tree runs.
- Get Keyword Suggestions Generates long-tail suggestions from a seed keyword using full-text discovery, with metrics on each. Complementary to Get Related Keywords rather than a substitute for it.
- Get Keywords For Site Discovers the keywords a domain already ranks for, with volume, difficulty, CPC, and intent. The workhorse read behind competitor gap analysis.
Backlinks
6- Get Backlink Summary Returns one aggregated view of a target's link profile: total backlinks, referring domains, IPs and subnets, broken links, spam score, and the breakdowns behind them. The snapshot before any deeper pull.
- List Backlinks Returns the individual links pointing at a target, with source URL, anchor text, dofollow status, link type, and the rank of both the linking page and its domain. The starting point for an outreach or disavow list.
- Get Referring Domains Returns the domains linking to a target, each with rank, spam score, contributed backlink count, first-seen date, and link-type and country breakdowns. Referring-domain count is the authority measure that survives scrutiny.
- Get Backlink Anchors Returns the anchor-text distribution of a link profile, with backlink count, referring domains, and spam score behind each anchor. This is the read that surfaces an over-optimized pattern before it becomes a penalty.
- Get Backlink Competitors Identifies the domains whose link profiles overlap most with a target, ranked by shared referring domains. Link-graph competitors are frequently not the same set as keyword competitors, which is exactly why this is worth running.
- Get Bulk Domain Ranks Returns the backlink rank for up to 1,000 domains, subdomains, or pages in one call. Used to triage a large prospect list by authority before anyone writes an email.
Site Audit
7- Start Site Audit Queues an asynchronous crawl and returns the task identifier every other Site Audit action needs. The crawl runs in the background for minutes to hours depending on page count, so the flow that starts it is not the flow that reads it.
- Get Site Audit Summary Returns a running audit's progress and aggregated findings: pages crawled and queued, detected CMS and server, and sitewide page metrics. The poll target while a crawl is in flight.
- Get Audited Pages Returns the individual pages a completed audit found, each with status code, title and meta description with lengths, heading structure, word counts, click depth, timing, and on-page score. The page-level detail behind a technical remediation list.
- Get Audited Page Links Returns the link graph the audit discovered, with source and destination, anchor text, attributes, dofollow status, internal or external direction, and destination status code. This is how broken internal links stop being a rumor.
- Get Duplicate Tags Groups audited pages by duplicated title or meta description, returning each duplicated value with every page using it. Among the highest-value fixes a crawl produces, and among the easiest to hand to a person as a work queue.
- Audit Single Page Audits one URL immediately and returns the same technical checks a full crawl produces for that page. Used when a flow needs a verdict on one page now rather than a crawl later.
- Run Lighthouse Audit Runs a Google Lighthouse audit against a URL and returns the full report: category scores plus every individual audit with its score and value. Categories can be restricted to keep the payload manageable.
Page Content
1- Parse Page Content Extracts a page's structured editorial content, separating header, main topics, secondary topics, and footer, and returning per-section headings, author, language, paragraph text, embedded links, and tables. Used to read a competitor's page as structure rather than as a wall of HTML.
Brand Citations
5- Search Content Citations Searches the citation index for pages mentioning a keyword or brand, returning each citing page with URL, domain, ranks, country, language, page type, and the surrounding snippet. The evidence layer under any brand-monitoring flow.
- Get Content Citation Summary Returns aggregate statistics for the pages citing a keyword rather than the pages themselves: total citations, overall rank, top citing domains, and the sentiment split. The read for a dashboard rather than a work queue.
- Get Citation Sentiment Analysis Splits a keyword's citations into sentiment buckets and returns a full aggregated summary for each, with an emotional breakdown. Used when the question is not how often a brand is mentioned but how.
- Get Citation Rating Distribution Buckets a keyword's citations by the normalized rating on the citing page, in tenths from 0 to 1, with an aggregated summary inside each bucket. Shows the shape of opinion, not just its average.
- Get Citation Phrase Trends Returns a time series of citation activity for a keyword by day, week, or month, with a full aggregated summary at every point. The read that tells you when something changed, which is usually the only question worth asking.
Technology and Domain Intelligence
5- Get Domain Technologies Returns the technology stack detected on a domain, grouped by area such as CMS, ecommerce, analytics, servers, and web development, with the domain's title, description, and rank. Used to qualify a prospect by what they actually run.
- Find Domains By Technology Finds sites using a given technology, category, or group, returning each domain with rank, title, description, country, languages, contact details, and full stack. The prospecting read behind competitive displacement lists.
- Find Domains By HTML Terms Finds sites whose homepage HTML contains specific markup, attributes, tags, or text, combined with AND across terms. Used to find fingerprints no technology detector has a label for.
- Get Technology Adoption Summary Returns aggregate adoption counts for a technology, category, or group without listing individual domains, broken down by country, interface language, content language, and associated keywords. Market sizing without pulling a million rows.
- Search Domains By WHOIS Data Queries a WHOIS database of over 130 million domains for registration and traffic profiles: registrar, creation, update and expiration dates, EPP status codes, and organic and paid keyword counts. Used to find expiring or newly registered domains in a niche.
Business Listings and Reviews
5- Search Business Listings Searches a database of over 150 million business listings scraped from Google Maps, with name, category, address, phone, website, coordinates, rating, and review count for each. The bulk read behind local market mapping.
- Get Google Business Profile Returns the live profile panel for a business: name, category, address, phone, website, coordinates, place and CID identifiers, claimed status, opening hours, and current operating state. Used to verify a location is represented correctly before anyone acts on it.
- Get Business Questions And Answers Returns the questions customers asked on a business's Google profile with the answers they received, including asker and answerer names and timestamps. Unanswered questions on a live profile are a customer-facing gap, and this is what finds them.
- Create Google Reviews Task Queues a review scrape and returns a task identifier. Google reviews are only available asynchronously, so this action returns no review text, only the handle to collect it with.
- Get Google Reviews Task Result Retrieves the reviews collected by a task, returning the business title and aggregate rating plus each review's text, star rating, timestamp, and reviewer name. The second half of the review pull.
Marketplaces
6- Search Amazon Products Returns live Amazon search results for a keyword, with organic and sponsored listings, title, ASIN, price range, rating and review count, and Amazon's Choice and Best Seller badges.
- Get Amazon Product By ASIN Returns the live product detail page for one ASIN, including title, brand, images, parent ASIN, the ASINs of every variant, and the marketplace URL. Used to enrich a catalog with what the marketplace actually shows.
- Get Amazon Product Sellers Returns every seller currently offering an ASIN, with seller name, who it ships from, current and regular price, condition, and seller rating. The read behind unauthorized-seller and price-erosion monitoring.
- Create Google Shopping Task Queues a Google Shopping search and returns a task identifier. Shopping data is asynchronous only, so this action returns no products.
- Get Google Shopping Task Result Retrieves the products a Shopping task collected, returning organic, paid, and carousel listings with title, price, seller, and product identifier.
- Get Pinterest Share Counts Returns how many times each of up to ten URLs has been pinned. Pin counts are a durable signal for visual and evergreen content, which makes this a useful tiebreaker when ranking pages worth refreshing.
App Stores
5- Search App Listings Searches a stored database of Google Play and Apple App Store listings, returning store identifier, title, icon, developer, price, rating and review count, and category for each app.
- Create App Info Task Queues a live fetch of one app's store page and returns a task identifier. App store data is asynchronous only.
- Create App Store Search Task Queues a Google Play keyword search and returns a task identifier, so the ranked list of apps for that query can be collected. The core action for app store optimization work.
- Create App Reviews Task Queues a Google Play review scrape for one app and returns a task identifier. Reviews can be limited to a single star rating, which makes a one-star pull a targeted complaint feed rather than a firehose.
- Get App Data Task Result Retrieves the result of any App Data task, selecting the same store and task type used to create it. The single collection point for all three app task types.
AI Search Visibility
5- Generate LLM Response Sends a prompt to ChatGPT, Claude, Gemini, or Perplexity and returns the model's answer with token counts, cost, any reasoning summary, and, when web search is enabled, the sources it cited. Used to observe what an assistant actually says about a market, on a schedule, with the citations attached.
- Search LLM Mentions Searches a corpus of real AI assistant answers for mentions of a brand, domain, or phrase, returning each matching prompt and answer with the sources the model cited. The evidence layer for the question of whether an assistant knows you exist.
- Get LLM Mention Metrics Returns aggregate visibility for a brand or domain across assistant answers: total mentions and combined AI search volume, broken down by location, language, platform, and the domains cited alongside it.
- Get Top Mentioned Brands Returns the brands assistants name most often within a topic or alongside a target, each with mention count, AI search volume, and its own breakdown. This is the competitive set as a model understands it, which is not always the one the market believes in.
- Get Top Mentioned Domains Returns the domains assistants cite most often within a topic, each with mention count, AI search volume, and breakdowns by location, language, and platform. The sources a model trusts are the pages worth getting onto.
Frequently Asked Questions
What can FlowRunner do with DataForSEO?
FlowRunner agents can run Google Organic Search, Google Maps Search, and Bing Organic Search in DataForSEO, plus 52 more actions.
Does connecting DataForSEO to FlowRunner require OAuth?
No. DataForSEO connects to FlowRunner with basic authentication (a username and password), no OAuth flow required.
Can DataForSEO trigger a FlowRunner workflow automatically?
DataForSEO doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Start building with DataForSEO
$100 in credits. No card required. Connect in minutes.