The decision, stated properly
Reach is the constraint on most go-to-market plans, and there are exactly two ways to hold it.
Build means you accumulate the attention yourself: a founder account, a company account, a newsletter, a podcast, search positions. Slow, permanent, and yours.
Borrow means you reach people through somebody who already holds their attention. Fast, rented, and available on a date.
Teams usually skip the decision entirely and arrive at a hybrid by accident, which is defensible as an outcome and expensive as a method. The two paths need different budgets, different hires and different patience, and picking one accidentally means resourcing it accidentally.
What building actually costs
The honest version of this cost is hard to find because most people describing it are selling the idea that it is cheap. It is not cheap; it is deferred.
The clearest benchmark in the market is the version built by people with no budget constraint. Andreessen Horowitz has assembled roughly a million followers on X, a daily newsletter at 250,000 subscribers and a podcast at around a million monthly downloads, with a marketing headcount that people tracking it put above 70.
Most firms run one or two. That is what the top of this path looks like: a media company staffed like a media company, inside an organisation that could afford to fund it for years before it paid.
Your version is a founder posting for twenty minutes a day and one content hire. That is a legitimate strategy and it is not the same strategy, and the timeline reflects the difference.
| Horizon | What building typically produces |
|---|---|
| Month 1–2 | Almost nothing measurable. Ranking is relative to your own baseline and you do not have one yet. |
| Month 2–4 | First returns on founder-led posting. Inbound starts mentioning posts by name. |
| Month 6–12 | Search positions arrive. The account can carry a small launch on its own. |
| Year 2+ | The audience is a real asset and the next launch starts from a higher floor. |
Two costs that stay off the spreadsheet. Attention decays without contact, so a quiet quarter lowers the baseline everything afterwards is measured against, which makes this a standing commitment rather than a project.
And it concentrates on one person: if all distribution runs through the founder, the company has a single point of failure who also has a company to run. Where founder-led breaks.
What borrowing actually costs
Borrowed reach is priced per campaign, arrives in days, and stops when you stop. Those three facts are the whole trade.
What you are buying is not impressions. It is a claim carried in somebody else's words, on their account, at their own reputational risk, which is a different object from the same claim on your channel and travels differently for that reason. The discount that applies to your own channel is the thing this route is buying past.
The honest costs.
- It does not accrue by itself. One campaign leaves a spike and some followers. A relationship maintained between launches leaves a bench you can call. That difference is a decision you have to make deliberately. A bench rather than a rental.
- You get who is available. Three to ten days gives the widest selection, and most launches go live inside five. A 48-hour turnaround buys whoever is free rather than whoever is right, and the difference shows up in the report.
- There is no rate card. The same creator is worth a different amount to different buyers depending on how much of their audience is your buyer, so the spread between quotes for one launch is often several times over. What a launch costs.
- It amplifies whatever it is given. Distribution is exactly as efficient at spreading a weak claim as a strong one, which makes the angle the highest-leverage unfunded part of the whole thing.
What each one leaves behind
The comparison that decides most of this is what is still there in six months.
| Build | Borrow | |
|---|---|---|
| Time to first real reach | Months | Days |
| Cost shape | Salary, continuous, returns late | Per campaign, immediate |
| Ceiling | Credibility, and your own posting cadence | Who is available and who fits |
| Left behind | An audience you can reach for free forever | Followers, coverage on other sites, a warmer baseline |
| Fails when | The founder stops, or nobody owns it | You buy reach for a claim nobody can argue with |
The row people miss is the fourth one on the borrow side. A launch that lands produces newsletter write-ups, threads and breakdowns on pages that keep existing, and those pages are what answer engines read months later when a buyer asks the category question.
Borrowed reach leaves an owned residue, and that residue is a large part of what you actually paid for.
When each is the right answer
Build first if you are defining a category and need the definition to be yours, your founder will genuinely post and enjoy arguing in the replies, you have eighteen months before it has to work, or your product is early enough that the audience should shape what you build.
Borrow first if you have a date, you are unknown in a category with incumbents, the launch has to produce pipeline this quarter, or your founder is not going to post and no amount of planning will change that.
The last one is worth being honest about early, because a distribution plan resting on a founder who will not do it is a plan that fails in month three.
Do neither if the product is not ready. Distribution accelerates the discovery that something is not good, and the people who make that discovery have public accounts and long memories.
The sequence that runs both
The answer for almost every company is both, and the value is in the order rather than the mix.
- Settle the claim. Costs nothing and decides everything downstream. One sentence a knowledgeable person could disagree with. Narrative design.
- Start the founder account now, whatever else you do. It is the cheapest asset on the list and the only one with a start date you control. Twenty minutes to write, some of the evening in the replies. The structure of a post.
- Borrow reach for the launches that have to land, during the same period. This is the part that produces the quarter while the slow asset is still slow.
- Keep the residue deliberately. The followers, the creator relationships, the write-ups, the report. Each borrowed launch should lower what the next one has to buy, and it only does that if somebody is treating it as accumulation rather than as a campaign.
- Let the ratio shift. Eighteen months in, a company doing this properly is buying less reach per launch for the same result, because the baseline moved. That shift is the whole return on the build path and it is invisible unless you measured the baseline at the start. What to measure.
Running them in isolation is the common failure. A founder posting into a void with no launch to point at, or a launch spike landing on an account with nothing to follow, both waste the half of the value that comes from the two touching.
Questions we get asked
How long does it take to build an audience worth launching into?
Founder-led posting typically shows returns in two to four months and reaches launch-carrying scale considerably later. For the size that makes a launch self-sufficient, look at the best-funded example in the market: Andreessen Horowitz reports roughly a million followers on X and a 250,000-subscriber newsletter, built with a marketing team reported above 70 people. That is the benchmark for the expensive path.
Is borrowing an audience just paying for ads with extra steps?
The difference is what the reader does with it. An ad is discounted on arrival because everyone knows who paid for the placement.
A creator writing in their own words puts their own reputation behind it, which is why the same claim carries further. Borrowed reach also leaves followers on your account and pages on other people's sites, and an ad leaves nothing.
Should we do both at once?
Yes, and in a specific order. Borrow the reach for the launches that have to land while the founder account posts through the same period.
The launches make the founder findable and the founder account lowers what the next launch has to buy. Running them in isolation wastes most of the compounding.