Marketing: Your CMO Agent
A complete marketing function — an expert CMO if you don't have one, a chief of staff if you do. It builds the pipeline model before it talks about channels.
Overview
Marketing advice has a particular failure mode: it arrives as a list of channels. Do content, do paid, do events, do ABM. It sounds like a plan, and it never tells you whether the numbers underneath it could possibly add up.
The CMO Agent starts at the other end. Ask it how to hit a revenue number and the first thing it does is build the model out loud — how many deals that target implies, how many opportunities that needs, how much pipeline that requires — and show you the arithmetic. Only then does it talk about channels, because only then can it tell you whether the channels are plausible.
It fills two very different seats:
- If you don't have a marketing leader, it is one. The plan, the budget, the ICP, the first channels, the board conversation. It teaches as it goes — names the term, then defines it in a clause — because it assumes no prior marketing training.
- If you *are* the CMO, it's your chief of staff. The VP Marketing who takes the CAC diagnosis, the board section and the brand-versus-demand argument off your desk, in practitioner language, without explaining what an MQL is.
Its defining behaviour is that it works as the front door to a team. When a question belongs to a specialist — content, creative, LinkedIn, research, analytics, enrichment, web, legal, finance — it hands the work to the Sapience agent built for it, and comes back with one answer rather than a pile of handoffs.
The Facts It Establishes Before It Answers
The agent won't answer a marketing question in the abstract, because the same question has different right answers depending on things you may not have mentioned. Before substantive analysis it establishes:
Fact | Why it changes the answer |
Stage and ARR | A first marketing hire and a scaled function are different problems |
What you sell, and to whom | The ICP is the input to almost everything downstream |
Average contract value | This is the big one — win rates and pipeline coverage both move with it |
Sales motion | Self-serve, sales-assisted and enterprise are not variations of one plan |
Channels running now | What already works is evidence, not a starting position to be replaced |
Budget and team | A budget "including salary" is mostly salary |
What CRM and martech you run | It asks. It will not assume you're on HubSpot or Salesforce |
It then shows you a short Facts Block — what it has established, what it has assumed (and what would change if the assumption is wrong), and what it still needs. Correct anything wrong and the analysis changes with it.
This is also why the agent remembers. Your company context persists between conversations.
What It Can Help With
Area | Examples |
Pipeline and demand | Working backwards from a revenue target, channel mix, ABM, product-led motions, what a plan actually requires |
Unit economics | CAC properly defined, blended vs paid vs new-logo, payback, LTV:CAC, why a number moved |
Brand and positioning | ICP, category and narrative, messaging, and the brand-versus-demand budget argument |
Measurement | What attribution can and cannot prove, incrementality and holdouts, the KPI chain, marketing ops |
The board | The marketing section, what belongs in it, and — pointedly — what doesn't |
The seat itself | Operating cadence, agency vs in-house, org design, a first-100-days plan |
Two things it does that are worth knowing about
It will question the "3× pipeline coverage, 20% win rate" rule of thumb. Both numbers move with deal size — larger deals win at materially lower rates and need more coverage to hit the same target — so a flat rule quoted without a contract value is the first thing the agent will push back on. It carries a working band by deal size for exactly this, and is careful to tell you what that band is: a planning aid, not a benchmark. It will say which parts trace to a published study and which parts it derived, and it will tell you plainly when a figure isn't solid enough to put on a board slide with a source line under it. Your own closed-won data replaces the band the moment you have two clean quarters of it.
It diagnoses a CAC rise in order, not as a list. Did spend rise, or did customer count fall? Did the mix shift toward expensive channels with every individual channel unchanged? Is it blended, paid or new-logo? Only then: saturation, funnel stage, measurement artefacts, timing. It then tells you which causes are benign arithmetic and which are the structural ones that should actually worry you.
It Is Honest About Attribution
Attribution is where marketing leaders lose credibility, in both directions — by over-claiming what it proves and by dismissing it entirely.
The agent's position: multi-touch attribution measures correlation with touchpoints, not causation. It systematically over-credits whatever is easiest to track and under-credits everything that happens before someone is in your dataset. So use it for directional channel decisions, use incrementality tests and holdouts when you need a causal claim, and take self-reported attribution seriously as a corrective.
It won't present an attribution model's output as proof that a channel caused revenue — which is exactly the overreach that hands a skeptical board the argument that marketing measurement is theatre.
Benchmarks Come With Their Sources Attached
There is no neutral, audited benchmark authority for B2B SaaS marketing. Essentially every widely-quoted figure comes from a vendor studying its own customer base, which skews toward that vendor's success stories.
The agent doesn't pretend otherwise. Figures are spoken with their source and vintage attached — "Benchmarkit's 2025 dataset, on fully-loaded CAC" rather than "the industry standard is…" — and it will tell you when a number came from a vendor's own customers. It also won't invent a figure it doesn't have.
The same discipline applies to the brand-versus-demand argument. The serious evidence there (the 60/40 split, share-of-voice models) is substantially B2C-derived, and its transfer to high-value B2B purchases is genuinely contested. The agent will say so rather than present 60/40 as established B2B fact — and will offer you a holdout test instead of a stronger opinion.
It Reaches For Specialists — On Purpose
Its instructions carry a table it consults before answering. When a question lands in a specialist's territory, it hands the work over and synthesises the result into its reply.
When your question involves | It reaches for |
Content strategy, editorial engines, SEO and AI-search audits, organic growth | Content & SEO Strategist |
Decks, graphics, brand collateral, web pages | Creative & Design Agent |
LinkedIn content calendars and actually publishing to LinkedIn | LinkedIn Agent |
Competitor, market and audience research sweeps | Web Research Agent |
Campaign, funnel, cohort or attribution analysis on your actual data | Data Analyst Agent |
Prospect and account enrichment, list building, ABM account data | Sales Development Agent |
Landing pages, microsites, interactive calculators | Coding Agent |
Agency and vendor contracts, sponsorship and influencer agreements | Contract Review Agent |
Ad-claims substantiation, trademark, advertising-law questions | Legal Research Assistant |
Marketing budget and ROI framing, unit economics, board finance | CFO Agent |
You don't arrange any of this. Ask in plain terms; what comes back is one CMO-grade answer in one voice, not a transcript of who was consulted.
The Content & SEO Strategist is the newest of these and was built specifically so this agent would have somewhere to send organic work. The CFO Agent is its sibling — marketing budget arguments are finance arguments, and the two are designed to hand work to each other.
How You Use It
- Install it from the Sapience AI Store, in the Sales & Marketing category.
- Tell it where you are — stage or ARR, what you sell and to whom, roughly what you spend, and whether you have a marketing team. Two sentences is enough.
- Ask the real question, with the actual numbers.
- Check the Facts Block and correct anything wrong before acting on the analysis.
- Ask for the artefact — the board section, the plan, the diagnosis write-up — rather than stopping at the explanation.
We're at $1.1M ARR selling a $28k compliance tool to financial services.
No marketing leader, $30k a month all-in, and the board wants $3M next
year. Where do I start?Good To Know
- It's a reasoning-heavy agent. It runs on Claude Opus 5 at high reasoning effort, so it will often think for a while before its first words. Longer waits on hard questions are expected, not a fault.
- It remembers your company between conversations, which is what stops the discovery step becoming a chore.
- It won't hand you internal machinery. You'll get the answer, not a report on which specialists were consulted.
- It is not advice. Marketing strategy guidance from an AI assistant — not legal, financial or investment advice. Benchmarks and market figures change constantly; verify anything you intend to act on, and consult qualified professionals for legal, financial and regulatory questions.
Summary
The CMO Agent is a complete marketing function: an expert CMO for a company that doesn't have one, and a chief of staff for the one who does. It builds the pipeline model before it talks about channels, defines CAC before it diagnoses it, attaches sources to every benchmark, and routes specialist work to the Sapience agent built for it — returning one answer rather than a set of handoffs. Install it from the AI Store under Sales & Marketing.