Relevance AI for Marketing Teams: Reports, Content, Lifecycle

Relevance AI suits marketing ops, lifecycle and content teams that want agents to draft the weekly campaign digest, turn one asset into channel posts and write lifecycle sequences inside HubSpot or Marketo, with a person approving before anything sends. The entry point is Pro at $19 a month billed annually or $29 a month billed monthly for the whole organization, with 2,500 Actions a month and two Build Users. Bulk scheduling, the analytics dashboard and End Users start on Team, at $234 a month billed annually or $349 a month billed monthly. Skip it if you want finished marketing agents without building anything. Those come through the sales team.

Monday's campaign digest, assembled from your own stack

The docs' campaign analytics guide describes an agent that reads HubSpot, Marketo, Google, Meta and LinkedIn ads, web analytics and the CRM, then posts a digest to Slack, Notion or email. The Marketplace template it points to, Email Marketing Strategist, analyzes campaign performance, works on subject lines, segments audiences and drafts copy. You clone it, then connect the channels it needs. A schedule trigger runs it weekly, and marketing agents can also raise alerts when click-through rates drop or send caps approach ^[1]. That covers the Monday report.

The Google Analytics connection is narrower than its name. It reaches GA4 only, not Universal Analytics, and offers four tool steps: Run Reports, Create Key Events, Create GA4 properties and a Google Analytics API Call ^[2]. Permissions follow Google's roles: Viewer to query, Editor to create key events, Administrator to create properties. One OAuth sign-in opens every GA4 property in that Google account. GA4 data can take 24 to 48 hours to finish processing, so a digest that needs exact figures should report on dates at least two days old. Point it at last week.

Relevance AI's sales-led Campaign Reporter shows what a mature version does: it runs on a Friday schedule, attributes channels to opened pipeline, posts a report with a budget recommendation to Slack and flags results built on small samples. A self-serve team can rebuild that pattern with a schedule, the HubSpot and GA4 steps and a Slack post at the end. Copy the small-sample flag too. A digest that treats twelve clicks as a trend misleads everyone reading it.

Turning one asset into a week of channel drafts

The Simple Content Repurposer template takes a YouTube link, a podcast episode or a blog URL and drafts LinkedIn and X posts in your voice. It has outside dependencies. YouTube sources need a supadata.ai API key, free for 100 requests a month, and podcasts need a paid spoken.md key. The docs' content-repurposing guide goes wider: one source asset becomes drafts for LinkedIn, X, email and ads, delivered to Notion or to schedulers such as Buffer or Hootsuite. The sales-led Content Repurposer queues its drafts in Notion for review rather than publishing them.

Visual work runs through Canva, connected by OAuth. Agents can create, edit and export designs, use the brand kit, manage folders, share and comment, and a Canva API step covers custom requests. As of November 2026, Relevance AI's docs still mark the Canva trigger as coming soon, so a finished design cannot start an agent yet. Notion connects through an internal integration token, and every page or database must be shared with that integration explicitly, since parent pages do not pass access down. Notion pages can also feed Knowledge, re-fetched every five minutes, and one knowledge base can sit behind several agents. That is where a brand-voice guide belongs.

For search work, the Google Search tool step returns organic results, People Also Ask and related searches for a chosen country and locale. On its blog, Relevance AI describes its own SEO setup: a monitor agent spots new or changed site templates and triggers a page-creator agent that researches search results, writes the page, formats it for Webflow CMS and submits it as a draft for human review. The agent drafts, a person publishes. Relevance AI says its social posting agent writes separate LinkedIn and X versions of each new blog post or changelog entry.

Lifecycle sequences and MQLs behind an approval gate

You give the lifecycle agent a segment, a lifecycle stage, a trigger and a goal, and it returns a drafted multi-step sequence with subject lines, body copy, send timing and a call to action for each step. Drafts land in HubSpot or Marketo for review, in a document for approval, or in your email service's queue, and the docs advise a Slack approval gate before anything sends automatically ^[3]. Approval is set per tool: Auto Run, Approval Required or Let Agent Decide. The Outlook docs suggest starting email tools in approval mode and switching to autopilot later.

HubSpot works differently from what many teams expect. An agent starts from HubSpot because you build a HubSpot workflow whose action sends a message to the connected Relevance AI agent, and that workflow can fire on a form fill, an email link click, a lead score being reached or a contact property update. Inside the agent, a HubSpot API Call step manages contacts, notes, tasks and engagements. So a new MQL entering HubSpot can open a qualification run, one of the three trigger patterns the marketing guide names next to a schedule and a campaign launch.

Marketo is thinner. Relevance AI built and supports the integration itself, and it consists of one tool step, Marketo API Call, which reaches any Marketo REST endpoint for leads, campaigns, email programs, statistics and lists. You connect it with a Client ID, a Client Secret and the REST API endpoint, and the step handles token refresh. The docs quote Marketo's standard limit of 100 calls per 20 seconds. Expect to describe endpoints rather than pick from menus. Gmail and Outlook cover outbound email directly, and our Relevance AI integrations guide walks through each app's setup.

Which marketing agents you build and which come through sales

Relevance AI's marketing function page lists MQL Qualifier, Inbound Engager, Lifecycle Marketer, Old Deal Re-engager, Content Repurposer, Campaign Reporter, SEO Brief Writer and Event Follow-up. These are sales-led engagements. The page leads with a demo booking, and Relevance AI's experts tune each agent. None of them is a Marketplace template you clone on Pro or Team. What you clone yourself is the Marketplace layer, including Email Marketing Strategist and Simple Content Repurposer, plus whatever Invent, the built-in AI builder, drafts from a plain description of your process.

The use-case pages still show patterns worth rebuilding. The Event Follow-up agent sends segmented emails after a Zoom webinar and scores hot leads in HubSpot with their question attached. The Lifecycle Marketer logs each send decision, with its reasoning, on the HubSpot contact record. The MQL Qualifier writes a score and flags to the lead record and pings the owner of a hot lead in Slack. Such agents can be chained on the Workforce canvas without code; our Relevance AI agents guide covers approvals and escalations, and Relevance AI for sales picks up where the MQL leaves marketing.

Pro or Team for a marketing department

Pro covers one or two builders whose agents run on schedules and triggers. Team becomes necessary for three marketing reasons. Bulk scheduling (Team and above) turns each row of a spreadsheet or knowledge table into one task, with working hours and pacing; separately, bulk run CSV uploads handle up to 10,000 rows. The analytics dashboard shows Actions and credits used by agent and workforce. End Users start there too. The marketing guide suggests evals to catch brand-voice drift, but agent evaluations are Enterprise-only ^[4]. On Pro or Team, the brand check stays human.

Volume decides the rest. Every Tool run is one Action whatever its number of steps, and failed runs and retries count too. A weekly digest running a handful of tools barely touches 2,500 Actions. Per-contact work is different: a lifecycle agent that runs three tools for each of 1,000 new contacts a month uses 3,000 Actions. By our arithmetic, Pro billed monthly plus a top-up pack of 1,000 Actions ($80 per pack) costs $109 that month for 3,500 Actions, and top-up Actions roll over to the next billing cycle.

That $109 counts Actions only. The model cost of drafting for 1,000 contacts draws on Vendor Credits: Pro includes 10,000 a month ($20 worth), and extra packs cost $20 per 10,000, so watch both meters in the pilot. If your model spend fits Pro's 10,000 Vendor Credits a month, Team becomes cheaper than Pro plus top-ups above about 6,500 Actions a month on monthly billing, or above about 62,000 a year (roughly 5,200 a month) on annual billing, by our arithmetic. Heavier model use lowers that crossover. Measure runs per contact in a pilot first.

Vendor Credits pay model and tool costs at wholesale with no markup, and your own model keys cut that spend but not the subscription or Actions. Outside services run on their own keys: supadata.ai, spoken.md and Firecrawl, which bills its credits itself. Every organization now needs a paid plan, and Relevance AI's pricing pages describe no free trial. Support on Pro and Team comes through the community forum, in-app chat tickets and email, staffed weekdays from 9am to 5pm Sydney time, with no weekend cover; engineers are on call around the clock for outages. For a US team, that window falls mostly outside office hours.

Who on a marketing team should build on Relevance AI

The clearest fit is a marketing ops or lifecycle lead who already maps processes in HubSpot or Marketo and wants agents that draft, score and report while a person approves the send. That lead gets the Workforce canvas, 2,000+ apps, GA4 and Canva connections and per-tool approval modes on Pro, then moves to Team once bulk sends or the analytics dashboard matter. A content team that only wants posts written faster gets less, since the repurposing template needs outside keys and, as of November 2026, the Canva trigger is still marked coming soon. Teams wanting the named agents built for them belong in the sales conversation.