The ground truth
What this actually looks like on a Tuesday.
Not the job description. Not what you told the board in your first week. What is actually happening in the building right now.
The real situation at $3M–$10M
—You are 1–3 people doing 8 jobs. There is no specialist. There is no agency budget. There is you, a generalist, and a shared Notion doc that nobody updates.
—The CEO still thinks they know marketing better than you do. They have an opinion about every email subject line. You are spending 20% of your time managing upwards.
—Sales has 2–4 reps doing outbound with inconsistent messaging, no enablement material that was built in the last 6 months, and a CRM that is 40% accurate on a good day.
—You inherited a website that describes the product the company used to be, not the company it is today. The ICP on the homepage is three pivots out of date.
—Pipeline is lumpy because it is almost entirely sourced from the founder's network. The question nobody is asking aloud: can we create pipeline without the founder in the room?
—The board has started asking about CAC. Nobody has clean data. The first three board slides about marketing are built on guesses dressed up as metrics.
"If your sales team cannot create an outbound conversation without the founder in the room, product-market fit has not been operationalised. It has been personalised."
This is the most important diagnostic question at this stage — and almost nobody is asking it directly. Founders close deals because they have conviction, depth, and relationships. That is not a repeatable go-to-market motion. It is a proof of concept. The Head of Marketing's job is to figure out whether the market actually wants what the company sells — or whether the company has been selling the founder.
The diagnostic — before anything else
The real PMF test — and why most companies fail it.
Most companies at $3M think they have product-market fit because they have customers. They do not. They have early traction. The difference is structural — and the Head of Marketing is the person best positioned to see it.
A false positive PMF is the most dangerous state a B2B startup can be in. You feel like you are winning. The unit economics are telling a completely different story. The gap shows up in marketing's first 90 days — if you know what to look for.
Signs you have early traction, not PMF
If two or more of these are true, stop scaling. Fix the foundation first.
✕Every deal was sourced by the founder. No rep has closed a deal independently in the last quarter.
✕The ICP is described differently by the CEO, the sales rep, and the website. Nobody has reconciled these three versions.
✕Customers "like" the product but would not be devastated if it disappeared tomorrow. Retention is acceptable. Expansion is rare. Referrals are non-existent.
✕Your best customers are outliers — they needed the product more than a typical buyer would. You have been convincing buyers, not attracting them.
✕Sales cycle length is increasing as you move away from the founder's network. New reps have lower win rates than the founder by more than 20 percentage points.
✕You are relying on discounts to close. Price sensitivity is high. Customers negotiate hard on every deal. This is not a pricing problem — it is a value problem.
Signs you have real, operationalisable PMF
The outbound test — can a rep replicate what the founder did?
✓A rep without founder involvement has closed at least 3 deals in the last 90 days. The pattern is replicable.
✓Churn is below 3% monthly. Customers who churned did so because of fit, not because they stopped needing the category.
✓At least 20–30% of new customers came from referral or word of mouth — without you asking them to refer.
✓When you raise prices on new customers, churn does not increase. This is the most reliable PMF signal most companies never test.
✓The market is pulling — inbound leads, unsolicited press mentions, prospects who already know what you do before the first call.
✓Net Revenue Retention is above 100%. Existing customers are buying more over time. You are not just filling a leaky bucket.
The outbound thesis
Why outbound is the PMF proof of concept — not a channel choice
For SLG and hybrid motion companies, outbound is not a growth tactic. It is the validation test. If a sales rep can pick up the phone, articulate the problem better than the buyer, generate a meeting, and close a deal — without the founder — you have PMF. Not before. PLG-native products (Slack, Figma, Notion) built virality into the product architecture from day one. They are the exception, not the model for most B2B SaaS. If your product requires a human to explain why it matters, that human needs to be able to do it without the founder in the room. Marketing's job is to make that conversation possible — not to have it instead of sales.
The motion split — this changes everything
SMB and enterprise are different jobs. Most companies are running both badly.
At $3M–$10M, many companies are trying to serve SMB and enterprise simultaneously with one marketing person, one message, and one motion. This is how you end up with a website that converts nobody and a sales team that is confused about who they are calling.
The SMB and enterprise motions require fundamentally different marketing architectures. Not different tactics — different operating models. Before you build anything, you need to decide which motion you are actually running, or — if you are genuinely running both — how to keep them from cannibalising each other.
SMB motion · ACV under $15K
Volume, velocity, self-serve signals
If your ACV is under $15K, your marketing job is demand generation at scale. You cannot afford a 6-month enterprise sales cycle per deal. The motion must be fast.
Build the inbound engine: SEO, content, product-led trials that create self-serve qualified demand for sales to convert
Nail the messaging — the homepage, the email sequence, the demo flow — so sales inherits buyers who already understand the value
Measure marketing by demo booking rate, activation rate, and time-to-first-value — not MQL volume
Build the ICP scoring model so sales can prioritise the signups most likely to convert — and stop wasting time on low-intent traffic
CAC payback target: under 6 months. If you're above 12 months at SMB ACV, the unit economics are broken and adding spend makes it worse
Key financial fluency: CAC by channel, free-to-paid conversion rate, activation rate, MoM MRR growth. These are the metrics that tell you if the SMB engine is working
Enterprise motion · ACV above $25K
ABM, air cover, buying committee
If your ACV is above $25K, your marketing job is not demand generation. It is account-based — making the right accounts know who you are before sales calls them.
Build the ABM list — every account that fits your ICP, built from closed-won data, not wishful thinking. This list becomes the single source of truth for marketing and sales
Set up custom audience ads on LinkedIn, Meta, and Google Display 30 days before sales outreach begins. The account should see your brand before the first cold email lands
Build sales enablement for multi-stakeholder buying committees — not a one-pager for the champion, but materials for the CFO, the IT lead, and the procurement team they will face
Create the category narrative — what is the problem this product solves and why does it matter now? Enterprise buyers need a reason to change, not just a reason to buy
Measure marketing by pipeline influenced, deal velocity, and win rate for marketing-touched accounts vs non-touched. Not impressions, not clicks
Key financial fluency: pipeline coverage ratio (you need 3–4×), sales cycle length by source, win rate by segment. Know these cold before any conversation with the CRO
If you are running both: treat them as two separate operating models that share a brand but almost nothing else. Different ICP definitions, different messaging, different channels, different success metrics. The fastest way to destroy both motions is to use the same playbook for both — which is what 80% of companies at this stage are doing.
The actual job
What marketing is building — and what it is not doing.
At $3M–$10M, the Head of Marketing is not the demand generation function. They are the system that makes demand generation possible. The distinction sounds semantic. It is structural.
Jason Lemkin at SaaStr puts it directly: at $1M–$10M ARR, the job is predictable demand generation — more leads each month. Brand comes after $10M. This is broadly right, but it misses the upstream work that makes demand generation possible. You cannot generate demand for a product whose ICP is wrong, whose messaging is confused, and whose sales team does not have the tools to convert what marketing sends them.
The five things marketing must own at this stage
1. ICP clarity — the foundation everything else is built on
Who is your real buyer — not who you think it is, but who has actually closed, stayed, and expanded?
Run the win/loss analysis. Take your last 20 closed-won deals. What do they share beyond obvious firmographics — the champion's title, their specific pain, what made them an internal advocate?
Define the anti-ICP. Which customers churned, stalled, or were painful to serve? These are the people you stop targeting. This is as valuable as knowing who to go after.
Get sales and the CEO to sign off on a single written ICP definition. Not a persona slide — a written document that everyone references when making targeting decisions.
2. Positioning and messaging architecture — before anyone writes a word of copy
Can you remove the logo from your homepage and still tell it apart from your three closest competitors? If not, you have a positioning problem, not a copy problem.
Build the messaging house: the core insight (why this problem matters now), the differentiated claim (what you do that no competitor can honestly say), and the proof (evidence that makes the claim credible).
Test the messaging with 5 people in your ICP who do not know your company. If they cannot articulate what you do and why it is different after reading the homepage, the copy is not the problem — the thinking is.
For enterprise: build the category narrative first. Why does the status quo fail? Why is now the moment to change? Enterprise buyers need a reason to move, not just a reason to buy your product.
3. Sales enablement that actually gets used
Most sales enablement at this stage lives in a Google Drive folder nobody opens. Build what sales actually needs: competitive battlecards that are honest about where you lose, objection handling scripts built from real call recordings, and case studies structured the way buyers think — not the way marketing writes.
For enterprise: build materials for every stakeholder in the buying committee — not just the champion. The CFO needs a different document from the IT lead. Procurement needs a different document from the VP who will sponsor the deal.
The test: ask three sales reps which piece of marketing content they used in the last deal they closed. If the answer is "nothing" or "I wrote my own," your enablement is not working.
4. Air cover for outbound — making the cold call warm
For enterprise and mid-market outbound: set up targeted advertising to your ABM account list 30 days before sales begins outreach. LinkedIn matched audiences, Google Display retargeting, Meta custom audiences. The goal is simple — when the SDR sends the first email, the account has already seen your brand three times. It is no longer cold.
For SMB: build the content that appears when your ICP searches for the problem they have, not the solution you sell. Top-of-funnel education content that builds trust before the intent is there.
Measure air cover not by impressions but by lift: do outbound sequences to accounts exposed to brand advertising perform better than sequences to non-exposed accounts? If yes, the air cover is working. If not, the message is wrong.
5. The measurement foundation — before the CFO asks for it
Build the attribution model before you need to defend it. Even an imperfect model told honestly is better than scrambling when finance asks how marketing contributed to revenue.
The metrics that matter at this stage: pipeline generated by source (not MQL volume), CAC by channel (not blended), pipeline coverage ratio, win rate for marketing-touched deals vs cold outbound. These are the numbers that make a CFO take marketing seriously.
Establish the shared definition of a qualified lead between marketing and sales before the quarter starts. The number one source of sales-marketing conflict is not strategy — it is that nobody agreed on what counts.
Capability gap — financial fluency at $3M–$10M
The numbers a Head of Marketing must know cold at this stage.
Not because finance is asking. Because the decisions you are making about channel mix, ICP targeting, and budget allocation are financial decisions — and if you are making them without understanding the unit economics underneath, you are guessing.
For SMB motion
CAC payback <6 monthsAt SMB ACV, if you are not recovering acquisition cost within 6 months, the unit economics do not work at scale. Every marketing channel decision should be evaluated against this benchmark.
Free-to-paid conversion rateIf you have a trial or freemium: what % convert to paid, in what timeframe, and which activation events predict conversion? This is where most SMB marketing spend is wasted — driving signups that were never going to convert.
CAC by channelNot blended CAC — broken out by organic, paid, outbound, referral. The blended number hides everything. Your organic CAC might be $200 while your paid CAC is $4,000. You need to know.
For enterprise motion
Pipeline coverage ratio: 3–4×The first metric the CRO will ask about. Total qualified pipeline ÷ revenue target for the quarter. Below 3× means you are going to miss. Marketing should be tracking this weekly, not quarterly.
Win rate by sourceDoes marketing-sourced pipeline close at a different rate than outbound-sourced? If marketing pipeline closes at 15% and outbound at 30%, you have a lead quality problem — not a demand gen success story.
Sales cycle length by sourceIf inbound leads take 30% longer to close than targeted outbound, that is a signal about intent quality. Know this number and have an explanation before the CRO brings it up.
Tools and leverage
What accelerates this work — AI, data, reading, and practice.
Claude is one tool in a system. Here is the full toolkit — what to use for which problem, and what to read to build the judgment underneath the tools.
AI — Claude
Win/loss pattern extraction
Paste 20 CRM deal notes (tagged W or L). Ask Claude to identify the top 3 patterns separating wins from losses, the ICP segments you win vs lose, and what the messaging implication is. This replaces a 2-week analysis project.
AI — Claude
ICP stress-test
Describe your current ICP hypothesis. Paste your closed-won data. Ask Claude to challenge the hypothesis — where does the data support it, where does it contradict it, and what refined segments should you test this quarter.
AI — Claude
Positioning audit
Paste your homepage copy and your top 3 competitors' homepages. Ask Claude to play a sceptical target buyer who has read all four. What impression do they have of each? Where do you sound identical to competitors?
AI — Perplexity
AEO and competitive intelligence
Ask Perplexity the questions your buyers are asking before they visit your site. "What is the best [your category] for [your ICP]?" — that answer is what your buyer is reading. If you are not in it, you do not exist in that moment.
Read
Founding Sales — Pete Kazanjy
The book that teaches marketing leaders how sales actually thinks about pipeline, qualification, and conversion. Read this before your next conversation with the sales lead. The language will change immediately.
Read
Lenny's Newsletter — PMF deep-dive
The most comprehensive research on how real companies found PMF — including how long it actually took Figma, Notion, and Slack, and what the signals looked like from the inside. Essential reading before you conclude your company has PMF.
Data
OpenView SaaS Benchmarks
Annual PLG and SLG benchmark data. If you are trying to make the case for a specific motion to your CEO or board, this is the data that anchors the conversation. Free-to-paid conversion benchmarks, CAC payback by ACV range, NRR by motion.
Practice
The 90-day marketing audit
In your first 90 days: run the PMF test, audit the ICP, stress-test the positioning, and map every piece of existing sales enablement against whether sales has actually used it. What you find will set your real agenda — not the one the CEO briefed you on.
What to read next
Not "more from this series." Routes to adjacent problems.
If the pipeline is the problem — where it comes from, why it is lumpy, and how to fix the architecture — go here. If the conversion is the problem — leads exist but deals are not closing — go there. Navigate by problem, not by topic.