Poldermarketing
How do I position an early-stage startup in a crowded market?
In a crowded market, buyers are already comparing options, so being only faster or cheaper than the incumbent is not a durable position. Name that path, then position against the alternative they use now, including doing nothing. Lead with one outcome a specific champion can check. Keep the public line loose until that buyer type returns on its own.
A crowded market is one path, not the only one
A crowded market is the Hair on Fire path: buyers already feel an urgent problem and are comparing products. A durable position there is a different customer experience. Being only faster or cheaper is a weak position on that path.
Sequoia’s Arc framework sorts companies by how the customer relates to the problem. Many founders assume they must be on Hair on Fire because they were told to listen to customers. Best-in-class on that path means different, not merely better, and the company still needs a great product and a great go-to-market in quick succession. Future Vision has the most ways to fail. Paths move. Do not treat one archetype as permanent.
| Path | Buyer state | What the position must do | Sign you picked the wrong path |
|---|---|---|---|
| Hair on Fire | Urgent, already comparing | A different experience versus the shortlist, plus a way to sell it quickly | Buyers are not comparing products |
| Hard Fact | Resigned to the status quo | A reason to change a habit, for a problem that matters enough | You talk like a feature bake-off and nobody is shopping |
| Future Vision | Unaware, or sure it is impossible | A believable next step, short of the whole ending | You sell the ending and have no next step a buyer can take now |
Sequoia’s cloud-security case was an agentless scan shown in a short demo. The HR case was a unified employee database. Rivals stitched separate datasets together, and the same layer then fed benefits, expenses, and devices.
Until the same buyers pull you, you hold a thesis
Until the same buyers pull you toward a segment, you hold a positioning thesis: guesses about alternatives, differentiation, value, ideal customers, and category. Across 16 in-house launches, that thesis was never exact.
April Dunford’s January 2026 note splits the work into prep, five steps that match the five components, and making the position real. Announce one industry only, and only that industry tries the product, so a miss teaches you nothing about anyone else. Keep the refusal list internal. Four decisions:
- Timing. Before launch you have a thesis, and you keep the public line loose. After buyers arrive, tighten only the segment their behavior has already pulled you toward.
- Audience. Customers hear what they get now. Investors hear the later path. Employees hear why someone would join: a product that lasts, room to learn, and a mission. One room cannot write all three.
- Object. With one product, product and company are the same position.
- Champion. That person builds the shortlist and recommends to the economic buyer. Miss them and you never meet the rest of the deal. Name a co-champion when two functions must both agree. When a third function can block the shortlist, name what that person must see, or you do not have a position yet. Security, legal, or procurement can be that block (Startupik). Do not invent a country rule the deals have not shown.
What a seed homepage should say is the customer sentence on the first screen.
The real alternative is life without your product
Start with what the buyer would do if you did not exist. Dunford’s quickstart then takes differentiated capabilities, value, target customers, and market category, in that order. Each depends on the one before it. Filling the blanks is pointless and potentially dangerous, because a product can sit in more than one category, against more than one alternative, for more than one buyer. Use the blanks only as a readout.
Doing nothing, in her quickstart, means the spreadsheet, the manual process, or an intern. Put a thin feature in a platform they already run on the shortlist only when it shows up in deals you lose. A rival who never appears in a deal is a phantom. Leave them off the list.
A desktop-database killer was barely remembered until buyers used it on mobile devices in the field. The frame that worked was an embeddable database for mobile devices. The same order turned a losing enterprise CRM pitch into CRM for investment banks, after a bank showed that relationship modeling mattered because it revealed who influences a live deal.
How to define the ICP is the “who cares a lot” step, written so you can refuse. Do it after the value is real.
Lead with the pain they already work around
The strongest pain to lead with is one a champion can replay from the last cycle, already works around, and can date.
- Recency. The last time: this week, this close, or this incident. “We want to grow” has no scene.
- Ownership. The person who can shortlist feels it.
- Workaround. Spreadsheet, freelancer, incumbent, or a manual patch.
- A date. Renewal, missed number, audit, launch, or a hire.
If two pass, lead with the one they would clear before the next trigger. Drop a pain the product does not change, and a pain whose owner cannot buy. Whether this is a go-to-market problem or a product problem is that fork.
A feature stays a feature until the alternative lacks it
A feature becomes the go-to-market line only when the current alternative lacks it, the champion performs the verb, you can show the result on their last incident, and the shortlist cannot add it without giving up how they already serve their current buyers (Beyond Ventures). If they can ship it as a feature, it is not the position.
The test is “so what”. “Anything you want” is not a value. Translate the capability, then a consequence the champion already tracks, then proof you can show this month. Show a workflow they can watch, not an unmeasured figure.
| Lead with it | Leave it off the first line |
|---|---|
| The alternative lacks it | The shortlist already has it |
| The champion performs the verb | A buyer you never meet |
| An outcome they already count | An adjective: faster, smarter, modern, AI-powered |
| Replay of their last incident | A case study you do not have |
Meet the features that shelf has taught the champion to expect. Leave those off the first line. Lead with the one the alternative lacks.
“AI-powered” stays on the first line only when the model does a job the alternative cannot, on a workflow the champion cares about, and a person still approves anything a customer will see. A wrapper on a general model is copyable. It is not the position.
Choose a frame the champion can picture
A position is easy to understand when, after one explanation, the prospect can say who it is for, what they would stop using, and why the switch is worth the hassle.
Dunford’s introduction treats positioning as context: who you compete with, which features are expected, and who it is for. Call it email and people expect spam filtering, storage, and a calendar. Call the overlap chat and they expect instant delivery and a sign a message was seen. She defines positioning as the market you intend to win and why you deserve to win it. A tagline, a brand story, and a vision come after.
Narrow a shelf the champion already shops until your value is the obvious one on it. Do not lock a segment you cannot reach (Startupik), or one that has no next segment which wants the same value. Invent a frame only when that shelf hides the value. The quickstart adds the CRM case: call it a CRM and a buyer assumes the category leader, a head of sales, and deal tracking. The same guide says the company which teaches a coined word often loses later to an entrant who arrives once the word is known.
One sentence has to survive a call: one champion, one job, one alternative, one proof, in their nouns. Put a competitor’s name in your place. If the sentence still fits, it is not differentiated.
Hold the public line until the same buyer returns
Use the same nouns on the homepage, in the first outbound line, and at the demo opening. Tighten only when the same champion type, the same alternative, and the same value repeat without prompting. One design partner is not that pattern.
Ask who it is for, what they would stop using, why the switch is worth the hassle, and what would make the claim believable. Open the line again when the champion type changes, the product changes, or replies stop matching the line.
A second channel will not repair a sentence people cannot repeat. A distribution wedge is not a second channel. It is an integration, a template, a community, or a step in a workflow the champion already opens, so the same position gets found again (Startupik).
The first go-to-market motion has to be one a founder can run. Testing positioning angles compares two frames on live replies. Turning the position into a campaign waits until a stranger has said the line back.
Leave it unlocked when the alternative is not in their words, buyers are not comparing, the only defense is speed or price, the person in pain cannot buy or you cannot reach them, you are about to coin a category, or the site line is the investor vision.
Before a campaign week, name the path, the champion, the workaround, the missing attribute, the consequence they already track, and proof you can show this month.
I split the customer sentence from the investor sentence
On a crowded market I write the customer sentence and the investor sentence as two documents, and I put one line in front of a champion before a campaign takes a week. I am Piet Baudoin, an AI-native growth marketer. I work alone, as the one-person practice Poldermarketing, fully remote, in Dutch and English. You can hire me freelance for a bounded project, or for a few days a week.
For a freshly funded startup I take marketing, AI, and automation as one go-to-market, from the first message to the first customers, without a separate specialist for each part. I am strong in AI, content, automation, and building workflows. I build and run the work. Google Ads and Meta Ads are newer for me. I can set them up and review them. I am the wrong hire when the core job is scaling a large media program.
The free growth scan shows how the current site reads. Positioning and messaging is where that sentence gets made, and how I work is the shape of an engagement.
Questions people ask
Should we invent a new category to stand out?
Invent a category only when the shelf buyers already use hides your value. Most early teams should narrow a known category until the difference is obvious inside it. [Dunford's quickstart](https://www.aprildunford.com/post/a-quickstart-guide-to-positioning) says the companies that do the teaching are often overtaken later by a company that arrives once the word is familiar. Her examples are Salesforce, which started as a niche inside CRM, and Qualtrics, spelled Qualtrix on her page, which started in survey software. Teach it before a sale.
How narrow should the first public line be?
Keep the internal thesis narrow, including who you will refuse, and keep the public line loose until the market pulls. Announce a single industry before anyone has bought, and only that industry shows up. A miss then tells you only that the industry was wrong. Tighten the homepage, the outbound line, and the demo opening together once the same champion type appears without you naming the segment. A single friendly account does not count as that pattern.
What if our only edge is speed or price?
If buyers are already comparing products, speed or price is a weak position. A durable advantage on that path is a different customer experience, and the company still needs a go-to-market that can move while the category is crowded. If you cannot name an experience the shortlist does not offer, do not scale a campaign. Go back to the alternative and find a capability it lacks, or admit you are not in a bake-off.
How do I pick the pain to lead with?
Require four things: a scene from the last working cycle, a champion who can put a product on a shortlist, a workaround they already run, and a date that makes waiting costly. When two pains pass, lead with the one they would clear before the next trigger. A loud complaint from someone who cannot buy, or a pain your product does not change, is not the lead. Wanting to grow fails the scene and the date.
Do investors and buyers get the same sentence?
Give them different sentences. Customer positioning covers now and the near term versus the alternative. Investor positioning is the path to a larger company later. Employee positioning is why someone would join. Choose the customer on purpose before any workshop. A room that tries to satisfy all three writes a soft sentence. Let the champion hear what they get now versus the alternative. Keep the later company in the deck.