Poldermarketing
How should founders run GTM before hiring a full sales and marketing team?
Run go-to-market yourself until someone else can follow a written path from a named account to a paid close. Narrow the buyer, stay on one motion, and log the ratio from your own list. Write stages, questions, and objections before you hire. Add a person only for the bottleneck that writing has exposed.
The founder sells until the path is writable
Run go-to-market until a second person can follow a written path from a named account to a paid close.
GTM Institute: you sell before a sales team exists, because only you can change the product, the price, and the pitch mid-call. Mercury: you cannot judge a miss you never ran. If the product is not built, GTM Institute’s order is problem interviews, a design-partner commitment, then the outcome by hand. Do not hire to discover that. Charge once there is something to charge for.
Write a buyer specific enough to refuse
Write company, person, and timing, GTMnow’s split from Pete Kazanjy’s workshop, plus the workaround, their words, why now, and who looks similar but closes badly. A live trigger makes a fitting account the one to contact this month. A budget freeze means not a prospect right now. A missing trigger is not a ban. Validate the ICP before outreach. Zeliq says stay where three happy customers already exist, and ignore the rest for two quarters.
GTMnow: write the story before the first note, as a spoken version, a short deck, the email, and a one-pager a buyer can forward. Read it to someone outside the market, then to someone in the profile. If the outsider cannot retell who it is for, do not send.
GTM Institute orders the first customers warm, then referred, then cold. Start from about thirty named people to land ten conversations. The cold slice of that ten is about twenty to thirty companies, one specific message each. Sending in volume before the note earns replies burns the accounts you need. If you cannot name 75 to 100 real companies, GTMnow says the profile is too narrow or still a theory. Do not send that whole list. This stage is personalized, not volume.
Stay on one motion until you own the ratio
Own one ratio before a second motion: named accounts, conversations, qualified next steps, and paid closes.
| What is true | Run this | Leave parked |
|---|---|---|
| Several people must agree | Founder outbound to a named list | Paid, a second segment, an SDR pod |
| The user can buy alone | One page, then calls | A field team |
| Buyers already gather in one room | Help there, then ask for a conversation | Cold notes to strangers |
| No names and no trigger | Problem interviews | Volume, scaled content, paid |
Mercury names sales-led, marketing-led, product-led, and partnership-led, and says test a few, then keep one or two. This page keeps one until that close has repeated. Pick by who decides, Zeliq’s buyer-versus-user question. Skip SDR hires, paid, and 30 SEO posts until the close repeats. Which experiments to run first is the kill rule.
Qualified means a current problem, a trigger, someone who can move a decision, and a dated next step. GTMnow gives Kazanjy’s sample as about 30 to 50 qualified prospects, about 10 to 20 converting. That is a sample size, not a weekly quota.
Ask what they do today, what it costs them, and what must be true to switch. Do not demo before priority. Kazanjy’s order is awareness, priority, preference, commercial agreement, then championship. Charge. GTM Institute treats free use as a different signal from payment. Concede with a longer pilot or later payment, and keep the price. On call one, name the user, the blocker, and whether the champion can get you in. One champion is not a close.
Give every stage an exit a stranger can use
A process is repeatable when a stranger can see what enters a stage, what leaves it, and why a deal died.
Stages: named, talked, qualified, evaluation, terms, won or lost. A loss exits in the buyer’s words.
Scott Travis wants it short: qualifiers a stranger can use, entry and exit, questions that sort intent from politeness, objections that moved a deal, and conversion by stage. GTM AI sets discovery at about eight to twelve questions you have already used. Objection replies are a frame, not a script. GTM Institute ends research when one page has held across at least three deals closed the same way. Before a full-time seller, GTMnow also wants good and bad discovery answers, a tested demo script, and a competitive sheet. GTM AI’s freelancer week tests that pack, not the one-pager alone.
Turn call notes into fields, then into a handoff
Every call leaves the same fields.
- Record with permission. Store their phrases.
- Log situation, trigger, objection, dated next step, owner, and won, lost, or no decision.
- Change one thing: pitch, price shape, or product.
- Within two weeks of a win or a loss, ask what almost flipped it. GTM AI frames that call as learning.
- Edit the one-page path on a set day.
- Before a full-time seller: about a week on the page, then three to five live calls. Pass if you recognize the discovery.
- Once a seller is seated, GTMnow reviews losses and repeated product gaps. If that stops, you left the learning as well as the sending.
From founder-led GTM to a system starts once that page is in use.
Hold calls, a list, and one public note in the same week
In one week, hold the calls, work one list, and publish one note cut from an objection you heard. Count proposals as their own step, and the days from first conversation to close. Change only the stage whose count moved. Content that supports outbound is that note.
Use LinkedIn against a named list
LinkedIn is an early channel when each action points at a named account.
Send one note about their situation and ask for a conversation. For an EU account, name yourself, honor an opt-out, and check that country’s rule before a commercial email. GTM Institute pairs a post about the problem with a message to people who engage and still fit. Ask for an introduction before a cold note. Approve every note until the transfer test passes. You choose the buyer and the send. Who runs lean outbound for seed comes after the list is real.
Leave personal outbound when the exits transfer
Stop all outbound yourself when someone else, under your rules, reaches a similar next step without you rewriting the note. Fatigue, spare cash, and pressure to hire a leader are not the test.
Published floors disagree. Use the passes, not one count. GTM Institute: no rep before ten or more personal closes. Startups.com: roughly 10 to 20 paying customers plus a written motion. Scott Travis: 15 to 20 deals on a consistent pattern, conversion stable, and pipeline is the bottleneck. GTM AI writes for founders who already have 15 to 20 deals to extract. That count is not the hire test. The freelancer row is.
| Pass | You still own |
|---|---|
| Three similar paid closes on one page: a research path exists | Every live conversation |
| GTM AI: recognizable discovery after about a week on the page | Price, odd deals, product promises |
| IGTMS, before a seller: closes without you, more than once, from more than one channel | Off-page deals |
| GTMnow, after the first seller: conversion tracks close to yours. If it does not, the document did not transfer | The weekly review |
Add one person for the constraint you can show
Add one person for one constraint you can point at, after that seat’s prerequisite is true.
| Constraint | Add | Already true | Not yet if |
|---|---|---|---|
| Only you can send, and replies come | A generalist on the list and one note | Profile, message, a reason per account | The note changes every send |
| Calls close, and time is the bottleneck | A full-cycle seller | Written exits, freelancer test passed | They must invent the pitch |
| The site disagrees with the calls | A marketer on the channel with a ratio | Buyer language from calls | Marketing in general |
| Stage names disagree | Part-time ops | Qualified is defined | A spreadsheet still matches |
| Buyers pay, then stall | Onboarding help | Manual steps are written | The buyer is still changing |
IGTMS skips a marketing machine while there is no validated message, and a fractional revenue chief while you are pre-revenue or have no product-market fit. A spreadsheet holds until about twenty active prospects. The funding round is not the test. UpliftGTM: a dedicated prospector fails before the profile and the offer are real, because the job is undefined and extra sending multiplies bad notes. Startups.com: full-cycle seller first, prospectors after that motion is real, a leader once several reps need a manager, often around five to eight. GTMnow: that seller runs your process, not a senior playbook. Fondo wants a clear site, lead capture, proof, and analytics in buyer phrases, then a clear customer, messaging that resonates, a path you can explain, channels that show promise, and growth work that no longer fits your week. Mercury: write whether the seat closes or builds the system. When to hire the first growth marketer is that gate.
Keep the motion unstaffed in four cases
Do not staff a team when you cannot name who you refuse, when no paid close has repeated, when the only wins are people you already know, or when a leader must discover the pitch. IGTMS warns that taking any yes pulls the roadmap apart. GTM Institute lets a model build the list, draft the first note, and summarize the call. You still run the conversation, and you still decide the one change to the product or the pitch.
What I take while the close is still yours
I work as one person, under the name Poldermarketing, fully remote, in Dutch and English. Hire me freelance, or for a few days a week. I am AI-native, and strong in AI, content, automation, and workflows. Google Ads and Meta Ads are newer for me: I set them up and review them, and I am the wrong person to scale a large media program. I build and run the work. The offer is an all-in GTM solution for a freshly funded startup anywhere, fully remote, from the first message to the first customers: marketing, AI, and automation, without a separate specialist for every part.
The free growth scan shows where the page and the calls disagree. See how I work and positioning and messaging.
Questions people ask
Should I hire a salesperson because I dislike selling?
No. Discomfort is not a stage signal. A seller cannot change the product, the price, or the offer mid-call, and cannot run a pitch that has not repeated. Hire when a second person can follow the written path and hit similar stage exits. If selling is the pain, cut the list to people with a live trigger and take those calls yourself. A job title does not invent the motion.
Do I need a CRM before the first customers?
No. Keep a spreadsheet while live prospects still fit on one view and you are the only person updating it. Move to one CRM when stage names, owners, or sources start to disagree. Write what qualified means before any route or score. A system that logs activity around an unclear buyer makes the mess faster.
What should a founder automate before hiring?
Automate account research, list cleanup, call summaries, reminders, and draft notes. You still choose the buyer, the message, the qualification, the price, and every note a stranger will read. Automate after a sentence has already produced conversations you can explain. A sequencer is not a reason to contact an account.
Do friends count toward a repeatable motion?
Use them only as a warm first wave, and only when they match the buyer you are testing. They will tolerate a rough product, which teaches you the setup. They are a weak read on who else will pay and why. Do not build the profile from friends alone. Charge. A free user answers whether the product is pleasant. A paying customer answers whether the problem is real.
How many acquisition channels should I run?
One, until that motion has closed more than once in the same way. A second channel may support it after that. Paid media and a large search program wait until the message is language buyers used on calls.