solving distribution

Most founders, me included, treat distribution as one big scary problem.

I need to grow. I need customers. I need marketing to work.

Framed that way, it’s unsolvable. There are too many variables, no clear first move, and when it doesn’t work you can’t even tell which part failed. You spend money, people sign up but don’t buy, and you’re left with a vague feeling of inadequacy and a pile of maybes:

• Maybe the ads are misleading
• Maybe they’re just looking for something free
• Maybe their pain isn’t high enough to pay to solve it
• Maybe the messaging set the wrong expectations
• Maybe they never experienced the value after signing up
• Maybe they didn’t understand what the product does after onboarding
• Maybe…

Every one of these might be true.

That’s exactly the problem: they all might be true, and staring at them as one undifferentiated pile, you have no way to tell which one is actually costing you the sale.

Worse, each maybe secretly belongs to a different part of the funnel.

• “Misleading ads” and “wrong expectations” are messaging problems.
• “Never experienced the value” is an activation problem.
• “Pain isn’t high enough” might mean you never had validation in the first place.

So “fix distribution” isn’t one job; it’s six different jobs wearing the same coat, and you’re trying to do all of them at once with no idea which one matters.

I worked this out because I was standing in exactly that spot. After many weekends, evenings, and vacations that went into hundreds of iterations of Mailover, I sat down and used it like a new user would, and everything worked. It saved me time, cut the noise, and I liked it. Half of the battle was done.

The other half was letting the world know about it. And I had no idea which part of that I was supposed to do first.

First, what distribution actually is

Before you break something apart, it helps to say what the thing is. Most people hear “distribution” and think marketing, or worse, ads. It’s both bigger and more specific than that.

Distribution is the machine that finds the person who has the problem where they already spend their time, says the thing that makes them care, clears everything standing between their pain and the moment your product proves itself, and turns that into growth that keeps compounding.

In plain words: distribution is getting the right product in front of the right person, making them instantly get why it matters, and making it easy enough to use that they stay.

A few other angles on the same thing, because each one catches something the others miss:

• Distribution is the systematic alignment of your product’s value with the habitats, habits, and psychology of one specific customer. It’s the work of removing every barrier between a person who has a problem and the product that solves it: the physical barriers, the mental ones, and the plain fact that the two have never crossed paths.
• Distribution is finding where your market’s attention is already concentrated, intercepting it with a relevant message, and redirecting it into an environment you control.
• Distribution is borrowing trust: transferring credibility from a platform, community, or creator your buyer already relies on to a product they’ve never tried.
• Distribution is articulating a problem so precisely to a specific segment that acting on your solution feels like the only logical next step.
• Distribution is machinery you can scale: channels, positioning, messaging, and mechanics that keep turning market interest into active product usage, at a cost that still leaves you a profit.
• Distribution is the end-to-end operational engine that locates an Ideal Customer Profile (ICP) in their natural habitat, engages them with resonant messaging, removes all friction between their pain and your product’s “Aha!” moment, and converts that value into sustainable, compounding growth.

Notice that none of those are “post more.” Every one of them is a fit problem: between a person, a place, a message, and a moment.

The four parts of a single attempt

Zoom into any one distribution attempt, a single ad, thread, cold email, or launch, and it always has the same four parts:

[ICP & segment] → [channel / habitat] → [resonant hook] → [desired action]

Who has the problem? → Where is their attention already? → What makes them care? → What do they do next?

Get all four right and the attempt works. Get any one wrong and the whole thing reads as noise. A great hook in the wrong room is invisible. The right room with a weak hook gets scrolled past. Both right with no clear next step is just entertainment.

That’s the shape of one attempt, and it’s what you’re building every time you sit down to write anything.

The first box is the one everybody fills in with a feeling. “Busy professionals.” “Small business owners.” Those describe a mood or a census category, and you can’t find either of them on a Monday morning.

Deep dive: finding your ICP, with the six questions that turn a vague audience into twenty names, worked end to end on Mailover.

What it doesn’t tell you is which part of your business is costing you the sale. Look back at the pile of maybes at the top: half of them aren’t in those four boxes at all. “They never experienced the value” happens long after the desired action. “Their pain isn’t high enough to pay” was decided long before the hook.

So the four parts tell you how to build one attempt. They don’t tell you what’s broken.

So stop treating it as one problem

It isn’t one. It’s a chain of smaller problems, each with its own question and its own answer. Once you break it down, you stop asking “why isn’t this working” and start asking the only question that’s actually solvable: “Which link is broken right now?”

Here’s how I think about that chain.

Two principles before the chain

1. Order by leverage. The links aren’t equal. Fixing an upstream link multiplies everything downstream of it. Reach is worthless if conversion is broken. Conversion is worthless if activation is broken. Activation is worthless if nobody stays (retention). The earlier the link, the more it’s worth fixing.

2. Earlier links gate later ones. This is the leaky-bucket rule. Pour water into a bucket with a hole and you don’t get a fuller bucket, you just lose water faster, and pay for the privilege. So you don’t earn the right to work on reach until the bucket actually holds water.

Spending on traffic before the funnel converts isn’t growth, it’s accelerated burn.

Keep both in mind, because the most common and most expensive mistake in distribution is working on a downstream link while an upstream one is still broken. I almost made it myself.

The chain

For each link: the question it answers, how you know it’s solved, and what failure looks like.

0. Validation: the gate before distribution

Not technically distribution, but nothing downstream matters without it. Do people actually have this problem, and does your solution solve it enough that they’ll pay?

• Signal it’s solved: real customer feedback plus real payments. Not “this is cool,” but money.
• Failure mode: you try to distribute your way out of a demand problem. Distribution amplifies demand; it can’t create it. If the underlying pull isn’t there, more ads just expose the gap faster.

Deep dive: solving validation, with the smoke test and the interview questions that separate politeness from intent.

1. Retention: does the bucket hold water?

Once a few people pay, the very next question isn’t “how do I get more”; it’s “do the ones I have stick around?” Talk to your first 50–100 users daily and iterate on the product based on their feedback until they love it enough to stay and keep paying.

• Signal it’s solved: churn flattening, cohorts retaining month over month, people renewing without being chased.
• Failure mode: people pay once and leave. Every dollar you spend on acquisition leaks straight back out the hole in the bottom. Growth here is a trap: it makes the leak bigger.

Deep dive: solving retention, with a calculator for the growth ceiling your churn sets.

2. Messaging: do the right people instantly get it?

For the people most likely to sign up and pay, is there a message they understand without explanation, that names the pain they already feel? Messaging is two things fused together: the wording and the audience assumption, who you’ve decided your buyer is. In the four parts above, this link is the who and the hook together, which is why you can’t fix it by tuning only one of them. If you’ve never written the who down as a sentence someone could check you on, start with finding your ICP before you touch a word of the copy.

• Signal it’s solved: people “get it” in one sentence, the right people show up, click-through holds.
• Failure mode: people arrive with the wrong expectations, or the right people scroll right past. A message that’s clear but aimed at the wrong audience fails the same way one aimed at the right audience but unclear does.

Deep dive: solving messaging, with a teardown template for your hero section.

Deep dive: positioning, the slot you’re claiming in the buyer’s head before you write a word of the copy, worked end to end on Mailover.

3. Activation: do sign-ups reach the aha moment?

The people who sign up, do they actually finish onboarding and feel the value, fast? This is the single most overlooked link, because it sits in the blind spot between marketing and product.

• Signal it’s solved: a high share of sign-ups complete the core action and hit the moment where the product proves itself.
• Failure mode: people sign up and never come back, because they never reached the point where they could see what it does. (My own number here was brutal: nearly half of sign-ups never even completed onboarding. They never gave the product a chance to work. You can’t convert someone who’s never seen the value.)

Deep dive: solving activation, with a friction-score audit for your onboarding flow.

4. Conversion: do activated users decide to pay?

These are all the moments where someone who has felt the value decides to buy, because they can picture their pain staying solved going forward.

• Signal it’s solved: a healthy free-to-paid or trial-to-paid rate.
• Failure mode: people love the demo but never pull out a card. Usually it’s pricing, friction at the moment of payment, or unclear ongoing value; they got value once but can’t imagine needing it every month.

Deep dive: solving conversion, with a decision engine for your pricing model and paywall.

5. Channel: where does your audience actually live?

Now, and only now, you go looking for the best place your audience hangs out, pays attention, and engages with what you put in front of them. This is the where of the four parts, and it’s the one everybody wants to start with, because it’s the easiest one to go shopping for.

• Signal it’s solved: one channel where the acquisition math works, where what you pay to get a customer is comfortably less than what that customer is worth.
• Failure mode: spraying across five channels at once, none of them profitable, and no way to tell whether the problem is the channel or everything upstream of it.

Deep dive: solving channel, with a matrix for which channels your price point can actually afford.

6. Reach: scale the channel that works

Once a channel converts profitably, the job is simple: do more of it. Spend more, reach more, get more leads, get more sales.

• Signal it’s solved: you can pour money in and it comes back out with a margin, predictably.
• Failure mode: scaling before the channel is proven, which is just scaling a loss. Reach amplifies whatever’s true. If the unit economics are negative, reach makes them more negative.

Deep dive: solving reach, with the creative engine that keeps a channel from fatiguing.

7. Referral: make growth compound

Once the whole machine works end to end, you get users to bring users. Incentives, loops, the things that make growth feed itself. This is trust transfer at its cheapest: the credibility comes from someone the buyer already knows, and you didn’t have to rent it.

• Signal it’s solved: organic and referred sign-ups become a rising share of total growth.
• Failure mode: bolting referral incentives onto a product people don’t love yet. You can’t pay people to recommend something they’re lukewarm about. Referral is a multiplier on love, and zero times anything is still zero.

Deep dive: solving referral, with a K-factor calculator for your viral loop.

The important caveat: it’s a chain, not an assembly line

If you take the order too literally, you’ll feel stuck waiting to “finish” a phase before you’re allowed to touch the next one. That’s not how it works. These links co-evolve.

You can’t finalize your messaging without knowing what the aha moment is, and the aha moment is an activation question. You can’t tell whether activation is broken until people arrive with the expectations your messaging set. Channel and messaging are tangled together too: B2B on LinkedIn and B2C on Meta aren’t the same words in a different room, they’re different messages entirely.

So treat the order as priority, not sequence. It tells you what matters most and what gates what. It does not mean you finish one link, lock it, and never look back. You loop through the whole chain constantly; the order just tells you where to spend your attention when you have to choose.

The tactics, and which link each one serves

There’s a whole library of distribution skills, and it’s worth knowing what’s in it. But the useful move is to file each one under the link it actually serves, because that’s what tells you when you’ve earned the right to use it.

Organic and content: the channels you own

• Audience building (build in public): documenting the journey on X, LinkedIn, or YouTube to earn trust before the product needs it.
• SEO and content loops: evergreen posts, comparison pages, and free tools that keep pulling traffic after you’ve stopped working on them.
• Newsletters and email lists: turning rented attention (followers on someone else’s platform) into owned attention (an address you keep), then writing the onboarding sequence that follows.
• Short-form video and copywriting: hooks written for the way TikTok, Shorts, and X actually spread things.

→ Mostly channel and reach. Building in public is the one exception worth noting: it doubles as a messaging lab, the cheapest place to find out which sentence lands before you pay to say it.

Product-led engines: growth built into the product

• Viral loops and referral mechanics: sharing designed into the product itself, like Dropbox’s extra storage, Typeform’s “powered by”, or Wordle’s shareable grid.
• Engineering as marketing: small free tools, calculators, generators, and databases that rank and travel on their own and feed the main product behind them.
• Community and ecosystem hijacking: a plugin in the Figma or Shopify store, an integration with Notion or Slack, a Chrome extension. Launching inside a distribution network someone else already built.

→ Referral for the loops, channel for the small tools and the ecosystems. Loops are the highest-leverage tactics on the list, and they have the strictest condition attached, because a loop only multiplies a product people already love.

Paid and partners: leveraged channels

• Creator, podcast, and newsletter partnerships: renting the attention of someone who already gathered your exact audience.
• Performance marketing: ads on Meta, Google, LinkedIn, or X, which is only a channel while what you pay to get a customer stays comfortably under what that customer is worth.
• Affiliate programs: commission systems that pay users, bloggers, and reviewers to sell for you.

→ Channel and reach, and the most dangerous family of the four. Paid is the one tactic that will happily keep turning money into traffic while an upstream link is broken. It doesn’t diagnose anything. It just spends more to arrive at the same failure.

Launches and platforms

• Launch tactics: Product Hunt, Hacker News, and Reddit. How to write a launch, and how to post in a community without getting flagged as a self-promoter.
• Platform algorithm mastery: how feeds on X, LinkedIn, and YouTube decide what to amplify.

→ A reach spike, not a channel. A launch happens once. A channel still works next month, and the month after. Treat one as the other and you’ll spend a year chasing spikes and calling it growth.

Now read the arrows again. Almost all of it answers where and how much: channel, reach, referral, links five through seven. Almost none of it tells you whether people stick, whether the message lands, whether sign-ups reach the value, or whether anyone pays.

That’s the trap. The loudest, most talked-about part of distribution is also the last part you earn. You’ll need most of these eventually, and they’re worth mastering. But a tactic pointed at a broken upstream link doesn’t fix the link. It just gets you to the same failure faster, and for more money.

How to actually use this

When growth isn’t working, don’t ask “why is my distribution broken.” That question has no answer. Instead, walk the chain from the top and find the highest link that isn’t solved yet. That link, the most upstream broken one, is your bottleneck. Everything below it is noise until you fix it.

More traffic into a leaky funnel just costs more money. The discipline is to resist optimizing the link you understand (usually reach: it’s the most visible) in favor of fixing the link that’s actually broken (usually something upstream and less obvious, like time-to-value).

And when you do get down to the channel and reach end, the four parts come back to do the actual work. Whatever tactic you pick, you’re still putting together who → where → hook → action. The chain tells you which problem you’re allowed to work on. The four parts are how you work on it once you’re there.

Knowing which link to fix doesn’t make the work feel any better. Distribution pays out on odds, and builders are trained on a game where good work produces the right answer every time, which is why so many of us stall here even with the map in hand.

Deep dive: why builders fail at distribution, on the switch from engineering absolutes to engineering probabilities.

That’s the whole point of breaking it down. Distribution stops being one unsolvable blob and becomes a list of small, ordered, answerable questions. You don’t have to solve distribution. You just have to find the one broken link and fix it. Then the next one.

One link at a time.