MCP App Store
by KeenEthics
Webmetic icon

Webmetic

By Webmetic GmbH

Get New App Alerts

Overview

Webmetic helps sales and marketing teams understand which companies visit websites connected to their account. Ask who visited, what they viewed, which accounts show buying signals, whether target accounts were active, or find contacts at a named visiting company. Website visitor data is read-only and identifies companies rather than people. Contact details are only revealed when the user explicitly asks for them.

Tools

check_accounts

ChatGPT
Target-account watchlist check (ABM): which of these named accounts visited the website in the date range — one call for the whole list. Match quality is shown per row ('exact' > 'prefix' > 'contains' > 'fuzzy'); ALWAYS verify the matched company name for non-exact matches — a name fragment can match a different company ('Bosch' can hit 'BoSch Data GmbH'). 'not seen' means no identified visit in the range OR a name mismatch: try the website domain, a company_id, or a wider range. Example: check_accounts(domain="example.com", companies=["Bosch", "kratzer.de", "Siemens AG"])

check_tracking_setup

ChatGPT
Load one of the account's own websites in a real browser and report whether the Webmetic snippet is there, carries this account's id, and runs before the consent banner. Use it when get_account_status shows a domain without visits, or when someone asks why no companies appear — it tells "not installed", "installed for another account" and "blocked by the consent banner" apart, which the visitor data cannot. The check loads the customer's own site, takes up to a minute, and runs one domain per call — never loop it over a list. It changes nothing, on the site or in Webmetic. Example: check_tracking_setup(domain="example.com")

find_contacts

ChatGPT
Decision makers at one identified visiting company — name, position, and which contact data is on file — the people to call after Webmetic identified the company. Not a general enrichment: the company must be in the visitor data. Contacts are limited to the visiting company's country (its seat in the visitor data); when nobody there matches, the result says so and shows contacts worldwide. Uses the account's contact credits: the search is currently free and cached per account; revealing an e-mail or phone number afterwards costs credits (reveal_contact). Returns named persons: call it only when the user asks for contacts or decision makers, one company per call, never in a loop over a visitor list. The account must have completed the contact setup in the Webmetic dashboard (target group + consent) — the error says so otherwise. Example: find_contacts(company="bosch.de", departments=["Marketing"])

get_account_status

ChatGPT
Readiness of the Webmetic account behind this connection: the tracked domains and whether each has ever recorded a visit, the plan and the Free cap, whether the contact setup is done, and the credit balance. Call this first when someone is setting Webmetic up, asks why they see no data, or before spending credits on contacts — it costs nothing and answers in one call what would otherwise be learned from a 402 or a 403. It does not say whether the snippet is live on the page right now; check_tracking_setup does that. Example: get_account_status()

get_activity_summary

ChatGPT
Aggregated overview of all visitor activity in a date range: totals, most active companies, top pages, traffic sources, industries, countries, devices. Use this for analytical questions ("which pages perform?", "where does traffic come from?", "how many companies visited?") — it computes everything server-side, never page through raw lists for aggregates. For period comparisons ("this week vs last"), call it twice with adjacent explicit ranges. Example: get_activity_summary(domain="example.com", from_date="-30 days")

get_company_activity

ChatGPT
Recent visit activity for one company: up to max_sessions sessions with time, duration, device, location, traffic source, and up to 15 page views per session (with time spent and scroll depth). Long results may be truncated to fit the response limit. Use after list_visiting_companies/get_company_profile to see what a specific company actually did on the site. Example: get_company_activity(domain="example.com", company="Bosch")

get_company_profile

ChatGPT
Firmographic profile of one company that visited the website: location, size, revenue, industry/NACE, contact data, registration, descriptions. The company must have visited within the date range (default: last 30 days). Example: get_company_profile(domain="example.com", company="Bosch")

get_visit_highlights

ChatGPT
Pre-filtered highlight lists: intensively engaged, returning, or first-time-ever visiting companies in a date range. Rows show the acquisition source instead of the industry (these feeds carry NACE codes, not the industry categories). 'new' is the only way to find first-EVER visitors (list_visiting_companies cannot tell a first visit from a repeat visit). Example: get_visit_highlights(domain="example.com", highlight="new", from_date="-30 days")

list_visiting_companies

ChatGPT
List the companies identified on the website in a date range, one compact row each. Call this FIRST for questions like "who visited my site?", "who viewed /pricing?", "which companies came via LinkedIn?". For aggregate questions (top pages, traffic sources, totals) prefer get_activity_summary instead of paging through this list. Drill down per company with get_company_profile / get_company_activity. A row named '[hidden] <company_id>' is a company hidden on the Free plan; upgrade to see it. Relevance = fit x activity, the number to rank companies by (0-100). Fit = how well the company matches the customer's ideal customer profile (0-100; '-' when the account has no profile for the domain, relevance then equals activity). Activity = the former lead score: how intensively the company visited (0-100). Example: list_visiting_companies(domain="example.com", visited_page="/pricing", sort="relevance")

reveal_contact

ChatGPT
Reveal the e-mail address (with LinkedIn profile) and/or the phone numbers of one contact from a find_contacts result. Spends the account's credits — e-mail 2, phone 8, both 10; data this account already revealed comes back for free. Reveal only what the user asked for, one contact per call. Example: reveal_contact(contact_id="v1.abc", reveal="email")

Capabilities

Writes

Example Prompts

Click any prompt to copy it.

Similar Apps