What should startup marketing look like before product-market fit?
Before product-market fit, marketing for a startup is mostly the founders finding users by hand and talking to them. Y Combinator's essential startup advice puts it in its pocket list: launch now, do things that don't scale, and don't scale your team or product until you have built something people want. It adds that growth is the result of a great product, not the precursor.
Y Combinator's Paul Graham, in the essay Do Things That Don't Scale (July 2013), is blunter about who does it. Founders can't hand early selling to a hire: "You have to do sales yourself initially." For B2B startups, the essay suggests sometimes picking a single user and acting like a consultant building just for them, and it names the big launch and the big-company partnership as tactics that usually don't work.
How do you know you are past it? Michael Seibel, writing in The Real Product Market Fit, says founders often think they have it when they don't, and describes the real thing as being overwhelmed with usage, with a frantic team dealing with ever-growing numbers of happy, loyal, ideally paying customers. Until then, Seibel says, stay lean and keep burn low. If you want a measured test instead of a feeling, our guide to running the product-market fit survey covers the 40% question.
| Before product-market fit | After product-market fit | |
|---|---|---|
| Goal | Learn which buyer has the urgent problem | Repeat what already brings customers in |
| Who markets | The founders, by hand | Founders plus the first specialist hire |
| Channels | Direct outreach, warm intros, communities where buyers ask questions | The one or two channels that produced paying customers, then tests of a third |
| Content | Founder posts about the problem and what you learn from calls | Content that answers buyers' questions at scale: guides, comparisons, case studies with permission |
| Spend | Time, not ad budget | Budget per channel, judged by cost per customer |
| What you measure | Conversations, problem confirmed, weekly active users, first payments | Weekly growth, customer acquisition cost, payback, retention |
Which marketing channel fits your startup's deal size?
Your price per customer decides how much founder time each deal can absorb, and that decides the channel. The table uses one target, $100,000 in annual revenue, and one example assumption we chose for illustration: you are willing to spend 10% of a customer's first-year revenue to win them, and you value founder time at $100 an hour.
| Annual price per customer | Customers for $100,000 a year | Founder hours you can afford per customer | First channel to try |
|---|---|---|---|
| $600 ($50 a month) | 100,000 / 600 = 167 | $60 / $100 = 0.6 hours | Communities, search, product sign-ups; outreach only for user interviews |
| $6,000 ($500 a month) | 100,000 / 6,000 = 17 | $600 / $100 = 6 hours | Founder outreach with a reason to write, plus founder posts |
| $24,000 | 100,000 / 24,000 = 5 | $2,400 / $100 = 24 hours | Researched outreach to named accounts, intros, small events |
| $100,000 | 1 | $10,000 / $100 = 100 hours | Account-by-account selling, pilots, references |
Customer counts are rounded up. Change the 10% and the hourly rate to your own numbers; the shape stays the same. At $50 a month, an hour of research per prospect loses money before you count reply rates. At $24,000 a year, skipping that hour is the mistake. Founder reports in our post on where B2B founders found their first 100 customers fit the same split: in one Hacker News thread, all seven B2B founders started by talking to people directly, and only the two low-priced products moved to SEO.
YC's advice also warns against chasing big deals with large companies early, saying they take too long and often fail, and it lists conferences as something to avoid unless they are the best way to get customers. So even at the top row, start with buyers who can say yes in weeks.
How should founders run content and outreach themselves?
Write to the few buyers who need it now, and post for the many who will need it later. The LinkedIn B2B Institute's 95-5 rule page (undated, checked 5 Nov 2026) says 95% of your potential buyers aren't ready to buy today. Its own conclusion is about advertising: invest in lead generation for the 5% in market, and invest far more in brand advertising that reaches the rest. An early startup rarely has a brand budget, so founder posts are the low-cost stand-in, and outreach covers the 5%.
The note gives a reason for writing, a narrow buyer, and an ask that costs 15 minutes. Pair it with two founder posts a week: one about the problem as buyers describe it on calls, one about something you learned building. Then check the rules for each channel before you automate any of it:
- LinkedIn. Its help page on prohibited software and extensions says it doesn't permit third-party software that scrapes or automates activity on its website, and cites User Agreement section 8.2 against bots that send messages, comment on, like or share posts. Our guide to what LinkedIn allows and what gets accounts restricted covers the detail.
- X. The X Developer Policy says services that post, follow or send DMs must follow X's Automation Rules, including getting explicit consent before sending automated replies or DMs, respecting opt-outs immediately, never bulk following, and never posting identical content across multiple accounts.
- Email in the US. The FTC's CAN-SPAM compliance guide says: "The law makes no exception for business-to-business email." Commercial emails need accurate headers, a non-deceptive subject line, your physical postal address and a working opt-out honoured within 10 business days; each violating email can cost up to $53,088. This isn't legal advice, and other countries differ; our comparison of cold email law under GDPR, UK PECR and CAN-SPAM covers them.
What does a first 90 days marketing plan look like?
This is our suggested plan for a B2B founder team, not a YC or Paul Graham method: one month each for conversations, hand-held users and choosing a channel.
Days 1 to 30: 40 conversations
Write one sentence on who you think has the problem. List 100 people who fit it, from your network first. Send 10 notes a working day like the one above, aiming for about 40 calls by day 30. Ask about the last time the problem happened and what it cost. Post twice a week about what you hear. No ads, no logo work.
Days 31 to 60: 5 to 10 users by hand
Narrow the list to the buyer type that answered fastest and cared most. Offer to set the product up for them yourself, or do the job by hand behind the product. Ask each user what they would lose if it disappeared. Start asking for payment by week 8, even a small amount.
Days 61 to 90: pick the repeatable channel
Look at where your active users came from. Keep the channel that produced most of them and drop the rest. Turn your best call notes into one useful public guide. Write down the message, list source and steps so someone else could run them.
The numbers in this plan are an example to adjust, not a benchmark: 10 notes a day for 20 working days is 200 notes, so 40 calls assumes one in five says yes. If your rate is lower, widen the list before changing the product. A sibling post on go-to-market strategy for startups covers pricing and positioning around this plan.
If you'd rather not run the list building by hand, Siela's Lead Finder suggests who to reach from your website, finds people with a reason to reach out now, and can be set to find leads only, which saves them to the CRM and sends nothing.
How do you measure marketing at an early-stage startup?
With one or two numbers, chosen to match your stage. YC's essential advice says a startup should choose the one or two key metrics it will use to measure success very early, then decide what to work on almost entirely by how a task moves them.
| Stage | Count weekly | Warning sign |
|---|---|---|
| Days 1 to 30 | Conversations held; how many described the problem without prompting | People are polite but no one has tried to fix it |
| Days 31 to 60 | Users active this week; users who paid | Use drops after you stop calling them |
| Days 61 to 90 and after | New paying customers by source; customer acquisition cost | One source brings sign-ups who never pay |
Customer acquisition cost is everything you spent on sales and marketing in a period divided by customers won in that period, so $7,500 spent to win 12 customers is $625 each. Count founder hours at a fair rate, or early outreach will look free. Our CAC calculator does the sum and adds payback time if you enter revenue and margin.
This week: write your one-sentence buyer, list 50 names, send the first 10 notes, and start a sheet with two columns: conversations held and problem confirmed.
Frequently asked questions
What are the most effective marketing strategies for startups?
Before product-market fit, the effective ones are manual: founders contacting likely buyers one by one, talking to users and helping early customers personally. After it, the effective strategy is whichever channel already produced your paying customers, scaled carefully. Broad launches, big partnerships and ads aimed at everyone tend to waste an early team's limited time.
What is a GTM strategy for a startup?
A go-to-market strategy is the plan for who you sell to, which problem you solve for them, what you charge, which channel reaches them and who does the selling. For an early B2B startup it can fit on one page, and it should change as calls show which buyer actually pays.
How much should an early-stage startup spend on marketing?
Before product-market fit, mostly founder time rather than money, because you are still learning who buys. Once a channel produces paying customers, set spend by what a customer is worth: work out customer acquisition cost per channel and keep it well below the revenue a customer brings in during the period you can afford to wait.
Should a B2B startup do content marketing or outreach first?
Outreach first in most cases, because it produces conversations within days, while content usually takes months to bring in buyers. Run both from the start in small doses: notes to buyers with a current reason to talk, and two founder posts a week drawn from those calls, so the content is ready when outreach shows what to write about.
When should a startup hire its first marketer?
After you can describe one channel that already brings customers, so the hire scales something proven instead of searching for it. Hiring before that tends to hand a stranger the founders' hardest job. Paul Graham's essay makes the same point about sales: founders sell first, and a hire replaces them later.
Sources
- Y Combinator: YC's essential startup advice (checked 5 Nov 2026)
- Paul Graham: Do Things That Don't Scale (July 2013)
- Michael Seibel: The Real Product Market Fit (checked 5 Nov 2026)
- LinkedIn B2B Institute: The 95-5 Rule (checked 5 Nov 2026)
- LinkedIn Help: Prohibited software and extensions (checked 5 Nov 2026)
- X Developer Policy (checked 5 Nov 2026)
- FTC: CAN-SPAM Act: A Compliance Guide for Business (checked 5 Nov 2026)