You have your first customers, revenue is rising, feedback is enthusiastic. Is the market pulling you, or are you pushing? The question is harder than it looks, because the two situations look alike for months: in both cases, things move.
Three reference texts let you handle it without slogans. Marc Andreessen's essay, The Only Thing That Matters, defines product/market fit. Paul Graham's essay, Do Things that Don't Scale (July 2013), explains why a founder must do everything by hand at the start. Finally, Y Combinator's essential startup advice connects the two. This article sets them side by side, proposes a stop test to tell pull from push, and says what it does not prove.
What Andreessen says: the market pulls the product
Marc Andreessen gives this definition: "Product/market fit means being in a good market with a product that can satisfy that market." He draws a consequence about the role of the market: in a great market, with lots of real potential customers, the market "pulls product out of the startup".
He also describes the two states. Fit is customers buying the product just as fast as you can make it, or usage growing just as fast as you can add more servers. Its opposite: customers who are not quite getting value out of the product, word of mouth that is not spreading, usage that is not growing that fast. He divides a startup's life into two periods, before product/market fit and after it, and writes that the only thing that matters is getting there.
His instruction for the "before" period is radical: do whatever is required, including changing out people, rewriting the product or moving into a different market.
What Graham says: at the start, you make it take off
Paul Graham writes that startups take off because their founders make them take off, and that the most common unscalable thing founders have to do at the start is to recruit users manually. He cites Stripe, whose founders would set the product up on the customer's own laptop ("give me your laptop"), and Airbnb, whose founders went door to door in New York recruiting users and helping hosts improve their listings.
Y Combinator's advice points the same way: get a first customer by any means necessary, even through manual work that could not be managed beyond ten customers. It adds that before you have made something your customers want, that is, found product market fit, it makes little sense to grow.
An apparent contradiction, and how to resolve it
Andreessen talks about a market that pulls, Graham about a founder who pushes. Both are compatible, as long as you do not judge the same moment. Pushing by hand is normal before fit. What separates pull from push is what happens when manual effort decreases.
The table below is our reading of the signs Andreessen describes, not a list taken verbatim from his essay.
| Question to ask | If it is push | If it is pull |
|---|---|---|
| What happens when you ease off for two weeks? | Sign-ups, usage and purchases fall back with your activity | Usage and purchases hold or keep growing |
| Where do new customers come from? | Almost all from your follow-ups, outreach and personal demos | A growing share arrives through referral or spontaneous demand |
| Does the product meet demand without you? | Each customer needs tailored hand-holding to get value | Customers get value on their own, and usage grows |
| What limits your growth? | Your time and your prospecting energy | Your capacity to deliver and to serve demand |
A stop test, to be done honestly
Our proposal, which follows from the two texts but is not taken from them, is to organize a limited voluntary pause: for two weeks, you cut part of your push actions (follow-ups, proactive demos, campaigns) without touching the product. You then note three things: new users, usage by existing customers and purchases.
Example: a founder selling a tool to small businesses closes ten sales a month thanks to phone calls. If they stop calling and sales fall to zero, they have measured push. If they still close four through referrals, they have measured the beginning of pull, to be confirmed.
What this test does not prove
- It does not replace a real judgment about the market. Andreessen makes the market the main variable: a good product in a bad market does not reach fit.
- Two weeks may be too short if your sales cycle is long. Adapt the duration to your cycle.
- It does not tell you what to do next. If it is push, Andreessen's answer is: do whatever is required to reach fit, which can go as far as changing market.
Further reading
- Validating product-market fit through customer feedback is about analyzing a segmented satisfaction survey. This article is not about surveys: it looks at what demand does when you stop pushing.
- Why hiring before PMF paralyzes your startup is about hiring. Here the subject is reading demand signals.
- How to Validate B2B Customer Demand Before Writing Code? is for a founder with no product yet. Here, you already have customers and want to know where their demand comes from.
Keep a record of the facts
A stop test is only worth something if you note the numbers before concluding. You can add your call notes and figures as sources of a Second Brain folder in Ember, then ask it to help you reread them before deciding. The decision stays yours. To discover the tool, visit Ember.
Sources
FAQ
Second Brain
Decide with project context
Ask a question and connect the answer to decisions already made in Ember.