Start here: find your next GTM decision

It's very normal for founders to get all the way to their launch with three crucial questions unresolved:

• whether the problem they're solving is felt keenly enough for someone to pay,
• whether the idea is differentiated enough to separate itself from the market, and
• whether anybody can understand that difference from the way it’s explained to them.

When one or all three fail, they launch to nobody; the launch still happens, but it becomes a rather private occasion.

I still use the candy, vitamin, or painkiller framework because it makes you decide what your product does for somebody - and fast. Does it kill your future customer's pain dead (a painkiller)? Is it a non-urgent improvement on their current situation (a vitamin)? Or is it something to enjoy but it wouldn't be missed if it went away tomorrow (a candy)?

This label doesn’t decide whether an idea will succeed - but it exposes what you're asking your future customers to care about, and how eager they'll be to give you their money.

You'll always find plenty of work to disguise the answer to this question: another feature, a launch calendar, a new homepage. Several social accounts and a sudden interest in SEO. Some of this may eventually help you reach people, but spending time on activities that don't directly address the core of your route to market is the wrong work.

This chapter gives you a starting point. Within it you'll be able to understand the problem-solving capabilities of your product, and you’ll leave with one GTM area to work on first, the evidence behind that decision, and the information that could change it.

Activity can hide the most important decision

The standard early-stage advice is heavy on action. Y CombinatorA startup accelerator and investor that runs an intensive program to help early companies improve their product, grow their users, and prepare to raise money. recommends launching early, talking directly to customers, and doing manual work before building systems for scale. Its advice includes an important qualification: the early product needs a “quantum of utility” for customers. There has to be enough value for a real person to experience and respond to. (Y Combinator, 25 September 2017)

This is good advice: there's definitely a case for putting something in front of people at the earliest opportunity, because buyer contact produces information that internal debate and research simply won't. If you have a usable product, and you're able to access a group of people who need it, you'll learn far more from them than another week iterating on a feature or a plan.

But your launch still needs a question behind it, and though this may sound harsh said out loud, “will anybody care about this?” combines the problem, buyer, promise, channel, price, and experience. If you're unsure of the answer, you need to make real human contact until you have concrete evidence to interpret.

A launch - an event you'll no doubt spend weeks anticipating and preparing for - is not just a one-and-done thing. It can expose your product to customers, but it can’t make an indifferent problem painful. More posts, more Product HuntA website where makers share new products and the community can discover, discuss, and vote for them. or Show HNA Hacker News format for sharing something you've made that other users can try, discuss, and give feedback on. upvotes may expand reach while leaving your core idea unintelligible. Traffic may reveal that people aren't taking the next step, but it doesn't tell you whether the problem is the buyer you're reaching, the price, the message, the product itself, the brand, or the page.

Even the word launch covers several different events. You may publish a URL, a discount, recruit people by hand, film some flashy product demos, hit up everyone you know to re-tweet you. But these events produce different evidence, and if you treat them as equivalent you'll be left with a pile of activity (perhaps some of it fruitful!), but little idea of what's been learned.

Paul GrahamA programmer, writer, investor, and co-founder of Y Combinator; his essay Do Things That Don't Scale is cited in this article.’s account of early user recruitment gets closer to the decision. He describes founders finding an initial set of users, observing who seems most enthusiastic, and seeking out more people like them. The manual work produces evidence about the customer as well as the product. (Paul Graham, July 2013) - It might feel like a thankless task at the time, but the most enthusiastic early customers will be your most valuable advocates and assets as time goes on.

What each signal tells you

A signal is something a potential customer says or does that helps you work out whether they care enough to act. Payment tells you something different from a compliment, and repeat use tells you something different again; that's why it helps to put these signals in order:

1. Payment: Obviously a tricky one to use if you're pre-revenue - if your product is in idea or MVPMinimum viable product: the smallest usable version of an idea that lets you learn whether it creates value for real people. stage, you won't be collecting cash. In these interim stages, can you look at other ways to judge this? A small pilot that's run manually or by hand, so you can see how the market responds to your proposition. Even a waitlist can help.

2. Repeat use: One of the biggest signifiers of product-market fitThe point at which a product solves a strong enough problem for a clear market, shown through demand and continued use. and product-customer fitHow well a product works for a particular customer or type of customer; they can use it, get value from it, and have a reason to return.. Are your users returning multiple times a week to use your product? If not, why not?

3. RetentionThe proportion of customers or users who continue using or paying for a product over a chosen period.: If you're a SMBSmall or medium-sized business. The exact cutoff varies, so this article uses the term only to distinguish smaller companies from large enterprises. SaaS product, and your monthly churn is less than 5%, you're doing something right.

4. Referral: Nothing says "I love this product" more than one of your customers recommending it to someone in their network. These are the advocates you need to treasure.

5. Usage: A signup gives you a starting number, but you need to know whether people reach and use the part of the product that does what you promised.

6. Stated intent: Someone saying they plan to try, buy, or recommend your product has made their interest more specific; you still need the action that follows.

7. Compliments: We'll talk about The Mom TestRob Fitzpatrick's method for customer conversations, built around past behavior, current problems, and money or effort already spent, so praise isn't mistaken for demand. later, but keep this in mind from the get-go. Someone saying that your product looks great always feels good - but it's a completely different sentiment to them actually parting with money to use it. One is good for your ego, the other is good for your business.

The same order can be used across different types of business, but the interpretation still needs context. Repeat use means something different for daily accounting software and annual tax software, and a long buying cycle delays payment, but a compliment remains a great thing to receive. It may show that somebody understands the idea - but it carries no proof that they’ll change their behavior or pay.

StrategyzerA company that publishes tools and methods for designing and testing business models and value propositions, including the assumptions-mapping method used in this article.’s assumptions-mapping method sorts beliefs by their importance and the evidence supporting them. The beliefs capable of sinking the idea, with the least evidence behind them, receive attention first. (David J. Bland, 4 August 2020)

That gives you an order for the work when you’re planning how to go to marketThe work of finding the people most likely to buy your product, explaining why it's useful, reaching them, and turning their interest into customers.; the important belief with the least customer evidence comes first.

The website promises something you're not

I recently reviewed a pre-launch product whose website opened by calling it “the X of Y.” The comparison was memorable. It also centered a part of the product that needed substantial social buy-in before it could grow.

That feature accounted for a small part of what the product did. Most of the product worked as an operating tool: it helped a person run an existing process. The opening line made the network-dependent feature sound like the product’s main job, so a prospective customer had to believe in the product and a future crowd before the rest of the value became visible. That's a much tougher job to sell - it feels like a vitamin, not a painkiller.

April DunfordA positioning author and consultant whose method begins with what customers would use or do if your product didn't exist.’s positioning method begins with the customer’s real alternatives, including their current manual process or decision to carry on as they’re doing. That comparison is more revealing than finding a fashionable company to place inside an “X of Y” line. (April Dunford, 28 December 2020)

How to find your starting point

The example above hid its main problem in the first line: the smallest part of the product had the biggest role. There wasn’t any customer response yet, so it was impossible to tell whether people wanted it, would pay for it, or would buy after seeing the website; on review, I could see that a future customer was being asked to understand the whole product through a feature that represented very little of its value. How the product was explained was the most important place to start.

Your own starting point will come from what has happened so far, and you can usually get there with four quick notes:

• What people outside your business have done: visited, joined a waitlist, subscribed, paid, returned, recommended you, signed up, replied, left, or decided against it.
• What you think may be true: the problem is urgent, one group will care more than another, or people will understand the product in the way you intend.
• What you’ve tried: building, posting, launching, changing the page, speaking to people, or selling by hand, along with what happened afterwards.
• What you have to work around: the time, money, access, technical limits, legal limits, or buying timescales that affect what you can do next.

The situations I see most often are below; the closest row to your own situation gives a solid starting point, and if several of these are true, the best one is the place that will make your next customer response easier to understand. In the example above, sending more people to a misleading explanation would have made it much harder to know whether they were rejecting the actual product, or the slim description that wasn't completely accurate.

If you don’t have enough customer response to choose a row, “I don’t know yet” is an acceptable answer because it tells you what to find out next. Strategyzer’s testing guidance uses more formal language; the TLDR is that a belief needs to be specific enough for real behavior to support or weaken it. “People will want this” gives you nowhere to begin, but a named group, problem, and expected action give you something you can learn from. (Tendayi Viki, 3 December 2019)

Your starting-point exercise

The Markdown starting-point worksheet records enough information to help you make one decision. Most answers can stay at one or two lines.

It asks for:

• The idea or product and what it helps somebody do
• Who you currently think it’s for
• How you explain it to a stranger
• What people have done, including payment, repeat use, referrals, signups, replies, and compliments
• What you’ve tried to get the product in front of people, along with what happened
• The limits you have to work around
• What you believe about the product or buyer but haven’t seen people demonstrate yet

The output names one place to start, explains what led to that answer, identifies what’s still missing, and gives you one small next step. If there isn’t enough information to choose, “I don’t know yet” is a valid result; the next step is to collect the smallest missing piece.

A prompt to find your starting point

Purpose: Identify the first GTM area that deserves attention.

When to use it: At idea, prototypeAn early representation of how a product could work, created so you can learn from it before the full product is built., pre-launch, or any later point when the next commercial move is unclear.

The prompt works in any AI tool. The bracketed text holds your context - fill this out - with each piece of evidence recorded separately. Direct your AI tool to store the output in Notion, Obsidian, or wherever you hold the context of your product.

I need to decide which go-to-market area to work on first.

My context
- Stage: [idea / prototype / live product / launched product]
- Product or idea: [what it is]
- Job it performs: [what it helps someone do]
- Who I think it is for: [the person or group most likely to care]
- Current promise: [the words a stranger sees or hears]
- Evidence collected: [payments, repeat use, retention, referrals, usage,
  stated intent, compliments, with dates or quantities where available]
- What I have tried: [activity and what happened afterwards]
- Limits I have to work around: [time, money, access, legal, technical, or buying-cycle limits]
- What I believe but haven’t seen yet: [list]

Your task
1. Separate my input into facts (what has happened), assumptions (what I believe
   without customer evidence yet), activity (what I have tried), and limits.
2. Assess the evidence in this order: payment, repeat use, retention, referral,
   usage, stated intent, compliments. Preserve the context of each signal.
3. Find the first important question I can’t yet answer, because later GTM
   decisions rely on it.
4. Recommend one place to start. You may choose any area, including validation
   (is the problem worth solving), first customer, ideal customer or ICP,
   narrowing the audience, positioning or how I explain the product, where I
   reach people, pricing, first use, repeat use, trust, or sales.
5. If the evidence can’t support one route, say "I don't know yet" and ask for
   the smallest missing piece of information.
6. If the information points in two different directions, give one other
   possible starting point and say what I would need to learn to choose.

Rules for the answer
- Do not invent customer behavior, demand, usage, payment, market facts, or
  results.
- Do not treat work I have done as proof that customers want the product.
- Do not force the answer to match a chapter that has already been published.
- Say what is still uncertain.
- Give me working notes I can use to make a decision.

Return exactly these headings
## What has happened (facts)
## What I believe (assumptions)
## What I have tried
## Limits I need to work within
## Where to start
## Why
## What I still need to know
## My next small step
## Another possible starting point (only if the information is mixed)
## Next chapter or area I need help with
## Next review

Good enough for now

The starting-point answer is usable when it names one GTM area, points to the evidence behind the choice, and records what could change it. It may be provisional - and that's fine, early evidence usually is.

Our next chapter is ValidationThe process of checking whether the problem, intended customer, and proposed product are supported by what people do; payment and repeat behavior carry more weight than praise.: if you can already show that people want what you're offering, you can skip ahead.

If you already have a live URL, Romy’s free First Read gives you an outside view of what the product appears to do, who it appears to serve, and what a stranger may struggle to understand. That can show you how a stranger reads your positioning. It can’t supply customer demand that hasn’t happened yet.

Sources

• Bland, David J. “How Assumptions Mapping Can Focus Your Teams on Running Experiments That Matter.” Strategyzer, 4 August 2020. Accessed 11 September 2026.
• Dunford, April. “Positioning and Competition.” 28 December 2020. Accessed 11 September 2026.
• Graham, Paul. “Do Things That Don’t Scale.” July 2013. Accessed 11 September 2026.
• Ralston, Geoff. “YC’s Essential Startup Advice.” Y Combinator, 25 September 2017. Accessed 11 September 2026.
• Romy. “First Read.” Accessed 11 September 2026.
• Viki, Tendayi. “Formulating Strong Hypotheses.” Strategyzer, 3 December 2019. Accessed 11 September 2026.