Metadata

Metadata is B2B ad campaign software where AI agents build, test and optimize LinkedIn, Meta and Google Ads campaigns for demand gen teams, sold by quote.

Metadata turns a marketer's outcome brief into running audiences, ads and budgets, then keeps adjusting them against pipeline rather than each network's own numbers. Metadata, Inc. owns and builds it; the company dates from 2015 and is based in San Francisco. It answers a familiar gap in B2B advertising: intent tools say which accounts may be in market, agencies add manual hours, and every ad channel grades itself. Metadata connects audience creation, campaign production and execution in one product, while the team keeps control of goals, budgets and approvals. Two generations share one login. Metadata Classic covers Demand Hub and Audience Only, with campaigns, audiences, converted leads, reports and integrations. Its experiments pair every audience with every ad and offer, and Budget Optimization moves spend toward the strongest performers on the metric the team picks. Auto-pause rules stop experiments on clicks, spend, cost per lead or cost per MQL, and Pacing flags early overspend and late underspend. The Daily Operations Digest reports both by email every day. Conversions return to Meta through its Conversions API, with LinkedIn's version in beta. MetadataONE, also in beta, puts an AI team behind a chat window that builds and launches campaigns, generates image creatives with editable layers and video ads, and reports spend and pipeline per channel. Metadata is built for B2B demand generation and account-based marketing teams whose paid budget runs mostly on LinkedIn and Meta, with Google Ads, Microsoft Advertising and Reddit alongside. It suits marketing departments that report on sourced pipeline and acquisition cost rather than on clicks. Enterprise teams that would otherwise hand campaign production to an agency fit it well, since the agents take on the building and the testing. It also runs product-led growth and retargeting programs. In practice the work starts with an account list from Salesforce or HubSpot, sharpened with intent data from Bombora or G2 and turned into audiences on each channel. Leads from those campaigns move through Marketo, Pardot or Eloqua to the opportunity in the CRM, which is where each experiment is finally judged. Marketers can also work with it inside ChatGPT, through their own assistant or in the MetadataONE app for iPhone, checking spend, leads and pipeline from connected ad accounts. The job it takes over is the cycle of rebuilding, retesting and reallocating paid campaigns that a demand generation team would otherwise run by hand.