Startup distribution

Something genuinely changed in the last couple of years. Building software went from a specialist skill with a long apprenticeship to something a determined person can do in a weekend with a coding agent and some patience.

What didn’t change is the number of hours in anyone’s day, or the number of new things they’re willing to try. The supply of software went up enormously. The supply of attention did not move at all.

The bottleneck moved and the advice didn’t

For twenty years, startup advice rested on the assumption that building was the expensive, risky part. “Ideas are cheap, execution is everything.” “Just ship.” Distribution came third, after the thing existed.

When building is cheap, shipping stops being an achievement and becomes the entry fee.

The result is a large and growing group in a specific position: a working, deployed, genuinely useful product, and no idea what to do on Monday morning. That’s not a skills failure. Nobody teaches distribution in a docs page.

Three things that are actually different now

“We built it” is no longer a story

Building in public worked partly because building was interesting. It’s much less interesting when everyone is doing it. What’s still interesting is the problem, who has it, what it costs them, what you learned about it that they don’t know.

The competition isn’t other startups

For most new products the real alternative is a spreadsheet, a group chat, or nobody bothering. That’s good news, habits are easier to displace than funded incumbents, but it changes the argument entirely. You’re not claiming to be better than a rival. You’re claiming this is worth changing a habit for, which is a much higher bar than most landing pages are written for.

Iteration speed is now a distribution advantage

If someone in a forum says “I’d use this if it did X”, you can plausibly reply the next day saying it does. Almost nobody exploits this. It’s the one genuine edge the new tooling hands you on the distribution side rather than the building side.

What hasn’t changed at all

• People only look for things when something has gone wrong.
• They describe their problem in their own words, not your category’s.
• They ignore anything that looks like an advertisement.
• They trust a specific, useful answer from a person more than anything else you can buy.

Which means the fundamentals are, annoyingly, exactly what they always were. The difference is how many people now need them, and how few arrived by a route that taught them.

The front door is moving

For twenty years the default way to find software was to search for it. A growing share of that now happens in a conversation with an assistant instead: somebody describes their problem in a chat window and asks what to use.

Nobody outside the model providers knows the real split, and anybody quoting you a precise figure is guessing. But the direction is not in much doubt, and it changes what being findable means.

• You are being described rather than ranked. An assistant summarises what your product is for. If your own pages are vague about who it serves and what it costs, the summary will be vague, and a vague summary loses to a specific one.
• Being mentioned elsewhere matters more, not less. Assistants draw on discussion, documentation and comparisons written by other people. The useful forum answer you left last month is now working two jobs.
• Plain, checkable facts win. What it does, who it is for, what it costs, what it does not do. This is the same discipline that makes a good landing page, which is convenient, because it means there is no separate programme of work here.

The distribution constraint, stated plainly

Attention is fixed. Software supply is rising fast. Therefore the expected attention per product is falling, and it will keep falling.

The consequence isn’t “market harder”. It’s that precision matters more every year. Broad, generic promotion was always weak; it’s now close to worthless. What still works is being specifically, verifiably useful to an identifiable person with a live problem, which doesn’t scale neatly, which is exactly why most people won’t do it.

What is actually available to a solo founder

Most channel advice is written for teams with a budget. Filtered down to what one person with a few hours a week can genuinely run, the list is short, and it is worth knowing the payback period before you pick.

• Being useful where the problem is discussed. Payback: days. Cost: your time. Ceiling: low, and it does not scale. This is almost always the right first channel and almost nobody wants it to be.
• Direct, individual outreach. Payback: days to weeks. Cost: your time, plus a real risk of doing it badly enough to damage your domain. Twenty by hand, never two hundred templated.
• Being the complement to something popular. Payback: weeks. An integration, a plugin, an adapter. You inherit somebody else’s audience and vocabulary.
• Directories and listing surfaces. Payback: immediate but small. Boring, deterministic, high completion rate, and genuinely worth doing on a day when you have twenty minutes and no energy.
• Launch platforms. Payback: one day, then nothing. A spike, not a channel, and you get roughly one per audience. Spend it when the page is ready, not before.
• Search and content. Payback: six to twelve months. A real channel, wrong for a this-month problem. The full argument is worth reading before you commit a quarter to it.
• Paid acquisition. Payback: requires a known conversion rate and a known customer value. You have neither, so this is not yet a channel, it is a way to buy a diagnosis you can get free.

Read that list as a sequence rather than a menu. The top two are available this week, and the ones at the bottom only become available after the top two have taught you something.

What to do with that

Three things follow from it.

Sequence over breadth. One channel understood properly beats five tested badly, because five bad tests produce five inconclusive results and no learning.

Diagnosis over planning. A plan written in week one assumes a bottleneck that has usually moved by week three. Marketing for founders has the diagnostic order.

Measurement, or none of it compounds. Marketing advice without outcomes never improves, including ours. If you don’t record what each action produced, you’ll be as uncertain in year two as you were in month one.

If you want the situational version rather than the argument, launched and nobody came and some users, no growth each start from a specific position. Or paste your URL into the Product Check, which is free, needs no account, and answers the only question this page cannot: which of these applies to you.