What is a go to market strategy, and how do I work out mine?

A go to market strategy is four decisions: who the product is for, what you say to them, what it costs, and the one channel you reach them through first. Write each as one sentence, then test them on real visitors. As an example only: an invoicing tool for freelance designers, "send an invoice in one minute", $10 a month, posted in one design community. Say 200 people from that community visit, 10 sign up and 2 pay. Those numbers tell you which decision to look at next. No official body defines the term, and for one person it is a page, not a forty page document.

Where the term comes from

No regulator, standards body or accounting rule defines "go to market", often shortened to GTM. It is a trade phrase for how a product gets from the person who built it to the people who pay for it, so the definition above is this page's own.

Paul Graham's essay Do Things that Don't Scale, written in July 2013 about the advice Y Combinator gives the startups it funds, treats the way you reach your first users as half of the idea itself.

The need to do something unscalably laborious to get started is so nearly universal that it might be a good idea to stop thinking of startup ideas as scalars. Instead we should try thinking of them as pairs of what you're going to build, plus the unscalable thing(s) you're going to do initially to get the company going.

Paul Graham, Do Things that Don't Scale, July 2013, under "Vector"

In plain words: what you build is one half, and how you get it to the first users is the other. That second half is your go to market, and the essay says the most common form of it is recruiting users by hand. The US Small Business Administration's page on planning a business never uses the phrase, but its lean plan asks the same questions under other names: customer segments, value proposition, channels and revenue streams.

What it is not, for one person

The same SBA page says a traditional business plan can be dozens of pages long, and that a lean one is typically only one page. For a product one person built, the go to market part is a few lines of that page. It is not:

  • A forty page document. You are its only reader, and you need four sentences you can test this month.
  • A list of every channel. It names the one to try first. How do I get customers for my app? says why.
  • A launch. The jobs around one launch day sit in a launch plan. The strategy is what is still true the day after.

How to work out yours, step by step

The formula in words: one kind of person, one sentence, one price, one channel, and the numbers you will count. Take them in this order.

  1. Who it is for. Name the one kind of person who needs the product most, and where they already talk about the problem. Y Combinator's essential startup advice says recruiting 10 customers who have a burning problem is much better than 1000 who have a passing annoyance. See ideal customer.
  2. What it says. One sentence in that person's own words: what the product does, who it is for and why it is different. That is positioning.
  3. What it costs. A price, shown on the page, that makes sense next to what the problem costs the buyer. See pricing.
  4. The one channel first. The place where those people already are, by a free method before a paid one. Distribution channels takes each from its own official rules.
  5. What you will count. Visits, signups and payments, each by source. The same advice says to choose, very early, the one or two key metrics you will use to measure success.

Then test it. An example in round numbers, made up to show the sums and not a benchmark:

DecisionThe example
Who it is forFreelance designers who bill by the hour.
What it saysSend an invoice in one minute.
What it costs$10 a month, shown on the page.
First channelOne design community where they already ask about getting paid.

Say 200 people visit from that community, 10 sign up and 2 pay. Ten signups from 200 visits is a conversion rate of 5%. Had nobody signed up, the first things to check would be who arrived and what the page said, not a second channel.

The mistakes a new founder makes

  • Every channel at once. Ancestorii, an earlier product of AskSami's founder, ran paid ads, a podcast sponsorship, SEO pages and cold messages at the same time, and had four signups after two months live. How do I get customers for my app? tells it.
  • Starting from the channel. Which site to post on is the fourth question, not the first.
  • Paying for traffic before the page works. A paid channel shows you faster whether strangers want the product. It does not make them want it. Should I run ads before anyone has paid? says when a test is worth running.

Where it matters

The channel decision is where your strategy meets rules written by somebody else. A community such as Reddit or Hacker News decides what you may post. A paid channel such as Google Ads or Meta Ads only pays for itself while customer acquisition cost stays below what a customer pays you. Whether the whole strategy is working is the question of product market fit.

How Sami uses it

AskSami writes an honest read of why the product does or does not sell. Its plan ranks up to ten ways to get customers, free methods before paid ones, one family of channels at a time. That ranking is the channel decision, made for your product.

Sami works from the numbers you give it and the records AskSami keeps. When it has no data on a channel it says so, and it never invents a figure. AskSami never posts anywhere for you, never spends your money and never writes to your code. You do those things yourself.

Questions people ask

How long should a go to market plan be for a solo founder?

One page. The US Small Business Administration says a lean business plan can take as little as one hour to make and is typically only one page.

Do I need a go to market strategy before I launch?

A short one, yes. Before the first post you need to know who the product is for, what the page says and what it costs, or the launch tells you nothing. Launch plan covers the days around the launch itself.

How many channels should my go to market strategy have?

One to start with: the place where the people who need the product already gather. If it fails, look at the product and the page before you pick another. How do I get customers for my app? sets out the checks in order.

How do I know if my go to market strategy is working?

Strangers from the one channel sign up, and some of them pay. No general figure counts as good, so count your own visits, signups and payments by source, and watch whether they move when you change one decision.

Read next

AskSami reads what you give it, all of it optional: your GitHub repository, up to 40 pages of your live site, and the screenshots, ad exports and payment exports. Before you sign up, Sami in the chat can search the web, open a few pages of your site and read a screenshot or a table you send, for that one answer. Once you sign up, your honest read and your ranked plan are free.

Sources

  1. Paul Graham, Do Things that Don't Scale, July 2013. Read on 8 October 2026.
  2. US Small Business Administration, Plan your business. Read on 8 October 2026.
  3. Y Combinator, YC's Essential Startup Advice, by Geoff Ralston, 25 September 2017. Read on 8 October 2026.

Checked on 17 October 2026. Every rule, price and figure on this page was read on the page it links to, on the day shown beside it above. They change, so open the source before you rely on one.