On Product Discovery · After Lenny's interview with Rob Snyder
PULL tells you what must be true. Scoring the struggle shows how strong.
Rob Snyder's PULL framework, from his book The Power of Pull and his conversation on Lenny's Podcast, gives founders and product people a plain test for whether anyone will buy what they are about to build. Here is how to pair it with Score the Struggle and progress interviews to find product-market fit before you write a line of code.
The Power of Pull
Rob Snyder · Basic Venture, July 2026. What You Need to Know About Customer Demand to Build a Successful Startup.
Discussed on Lenny's Podcast: "How to find product-market fit (but, actually)."
recognize a buyer who needs a better option now.
compare opportunities: the stake, the urgency, and what people have already done.
Customer Progress and No-Decision interviews show where PULL already exists, and where it breaks, before you build anything.
Five takeaways
- Pull belongs to a person, not a company. You sell to the head of legal at SpaceX, not to SpaceX.
- Score urgency and evidence separately. Urgency is what people describe. Evidence is what they had already done before you arrived.
- A tolerated struggle is not a dead one. Real but survivable often means Latent demand waiting for a trigger.
- Interview people doing the job, not just your customers. Switchers show where pull exists. Non-switchers show where it breaks.
- Count actions, not reactions. "That's the future of our industry" is a polite exit. "How do we get started?" is pull.
00 · The problem
Why zero to one is so hard
Most founders follow the same playbook: find a hair-on-fire problem, build a differentiated solution, and sign design partners to prove it works. Then the same thing happens.
- A hair-on-fire problem
- A differentiated solution that provides real value
- Design partners using the product every day
Prospects agreed. They didn't buy.
They confirm the problem and praise the product. So founders sharpen the demo and the pitch. Design partners use it every day, give great feedback, and never convert.
We ran into the same zero to one traps Rob describes: building based on what prospects said they wanted rather than what actually got them to say yes.A member of Lenny's community, read on Lenny's Podcast
The trap is that validation and buying are different questions. People can confirm a problem is real, praise the solution, and use it for free, and still have no reason to drop their current priorities to pay for it. Three things make that gap hard to see:
Snyder lived all three. His first startup checked every box, then spent two years pushing a product the market didn't want. The turn came when a Dunkin' franchisee who didn't want the product offered to pay for something else. That pivot took the company to $4 million in revenue.
Marc Andreessen described product-market fit as the moment "the market pulls product out of the startup." PULL makes that testable for one buyer: a specific person stuck on something they have to get done now, with no option good enough. Scoring the struggle adds how strong it is, and what people have already done about it.
01 · The framework
What PULL says
Snyder, author of The Power of Pull, frames pull as four things true of one buyer at the same time. When all four hold, the buyer is a coiled spring. It would be weird for them not to buy.
There is a project on their to-do list.
It has to get done right now.
They have looked into their options.
Those options fall short for a critical reason.
This is the hair on fire problem very concretely defined.Lenny Rachitsky on Project and Unavoidable, Lenny's Podcast
You hear complaints, but nobody buys.
You hear "thanks, we're handling it ourselves."
Pull is discovered, not created. It sits on someone's to-do list before you show up, so the work is finding it, not persuading people to feel it.
A note on the word "pull"
A project they must get done now, with no option good enough. It exists before your product shows up.
One of the six forces: the attraction of a specific new solution, weighed against the buyer's Anxieties about it. It only exists once a solution is on the table.
In this article, pull means the first: the buyer's demand, not a product's appeal. In my own JTBD work, I capture the buyer's full state in a Desire-for-Progress Statement, six forces in one sentence. Snyder's Project and Unavoidable map to its Trigger and Push. His Look and Lacking explain why the status quo fails. The statement adds what PULL leaves out: the Pull toward a specific solution, the Desire behind it, the Anxiety that can stall it, and the outcome the buyer wants to Avoid.
02 · The instrument
What struggle scoring adds
PULL already asks the right behavioral question: has this person looked into alternatives, and are they trying to buy? Struggle scoring makes the evidence behind that answer explicit, so you can compare one opportunity against another before you build.
The scoring key
Severe, but caught before people have acted.
The stakes and the proof together.
Unlikely to create demand.
Real but survivable. People live with their workaround.
Urgency: what they described
- Disruptive, not just annoying
- Necessary to resolve, not optional
- Hard to postpone
- Significant consequence if ignored
When "hard to postpone" or "significant consequence" is checked, capture the Impact of Delay: what happens if they wait, in time, money, or identity terms, and when it must be addressed.
Evidence: what they had already done
- Nothing yet
- Actively searching
- Built a workaround, or committed time or money
- Paid for, switched to, or seriously attempted a replacement
The interactive rubric scores one struggle in about ten minutes, and shows Gong go from a 0 to a 3.
Three levels of evidence
Each level supports a different conclusion, so score the highest thing they have actually done.
Actively searching for a better way: alternatives evaluated, demos taken, peers asked.
Built a workaround, or committed additional time or money: manual notes, a spreadsheet, a contractor.
Paid for, switched to, or seriously attempted a replacement.
That is what the PULL conversation tests. Lean-ins show pull exists; the score measures what people had already done before you arrived.
The stake explains the urgency
Naming what is primarily at stake (time, money, risk, or identity) tells you where "unavoidable" could come from. Any stake can become urgent: time through a deadline, money through a cash crunch or a lost contract. Any stake can also be tolerated for years. Three questions tell you which, and together they make up the Impact of Delay:
What happens if they wait?
When must it be addressed?
Who answers for it, and how are they seen if it slips?
State what happens in time, money, or identity terms. Answer when in their planning horizon: this month, this quarter, this half, or this fiscal year. Who bears it joins accountability and identity: the person on the hook for the outcome is also the one whose standing is at stake, and usually the one who can buy.
03 · The case
Same job, two scores: Jump
Snyder's clearest example is Jump, an AI notetaker for financial advisors. The same job scores a 2 for one group and a 0 for another. The difference is the stake.
When client meetings pile up, I realized writing notes after each one was no longer working well enough, because it takes time I'd rather spend selling.
When every client meeting has to be documented in the CRM for compliance, I realized writing notes by hand was no longer working well enough, because it eats hours and a gap could put me at regulatory risk.
The generic notetakers fell short for a specific reason: advisors could not tell whether they were compliant. That is PULL's "Lacking for a critical reason," and it is also an Anxiety. The tools may have worked fine. Advisors just could not trust them with a regulated record.
A 2 is the rubric working as designed: high urgency, caught before advisors had built or bought their way around it. The before score is reconstructed from what Snyder describes in the episode. What turned that early signal into demand was the conversation itself, when advisors started pulling the product out of Jump's hands. The same struggle, scored today, shows what followed.
High urgency. Level 1 evidence: they had looked at generic AI notetakers and ruled them out over compliance.
Level 3 evidence: advisors have switched to a compliant AI notetaker built for their CRM and regulators.
What followed: per Jump, more than 50,000 advisors now use it, at firms including LPL Financial, Northwestern Mutual, Allianz, and John Hancock. The after score does not validate the before score in hindsight. It shows the early signal was real.
Taking the product to insurance agents is a Door One question from Score the Struggle, Not the Solution: the same job, a different executor. Scoring it first shows why the move would not hold. The job carries over. The stake that made it unavoidable does not.
04 · More from the episode
Snyder's other examples, scored
Read through the rubric, most of Snyder's stories line up with their scores. These are illustrative reconstructions from what he describes, not predictions made in advance.
| Example | Stake | Urgency | Evidence before the call | Score | What happened |
|---|---|---|---|---|---|
| Pharma exec, market research too slow | Time | High | Level 2: committed money, already paying consultants for market research | 3 | "If you can do that, I will pay you right now" |
| Government buyer, applicant tracking | Time | Low | None | 0 | Lukewarm call |
| Same buyer, job descriptions (Hail Mary) | Time | High | Not described before the call | 2 | Offered to pay on the call. Became the first contract |
| Snyder's HR software, better UX | Time, felt by users but not the buyer | None for the buyer | None | 0 | Failed |
| A 100x better home printer | None | None | None | 0 | "I won't buy it because I'm not blocked" |
The pharma exec's money already going to consultants mattered more than the excitement in the room, and it gave the founder a natural price anchor too. The HR software failed on a question the rubric forces you to ask: whose struggle is it, and can that person buy? The printer is the thesis in one line. A solution can be 100 times better and still go unbought, because the struggle underneath it scores zero.
05 · Discovery
Finding PULL before you build
Snyder's advice for finding pull is to get your hands dirty. About 90% of the startups he has helped found their idea by going out and doing the job firsthand, not by brainstorming. Immersion lets you feel the struggle. Interviews let you see whether it repeats and reconstruct the evidence behind it.
The people to interview are not only your customers. They are people doing the job today, whatever they use: a competitor, a contractor, a spreadsheet, or nothing at all.
Where each PULL condition surfaces in a switching timeline
The struggling moment, and what was at stake.
When tolerating it stopped being acceptable.
What they actually investigated.
Why each option fell short, and what they settled for.
Customer Progress Interviews
A Switching Interview paired with a Core Jobs Interview, with people who recently made progress on the job using anything.
- Reconstructs all four PULL conditions
- Grounds the stake and the trigger in what happened
No-Decision Interviews
People who had the struggle, looked, and stayed with what they had, including those who simply tolerate the workaround.
- A project with no trigger yet: Latent demand
- Urgency blocked by anxiety or habit: what to de-risk
- Options judged good enough: the bar to clear
Where PULL sits on the demand spectrum
On the four stages of demand, PULL marks the move from Latent to Emerging.
Project present. Nothing unavoidable. Workarounds tolerated.
All four PULL conditions. Real switches start.
Size and score the repeating job.
Optimize the job at volume.
The signals for leaving Latent read like a PULL checklist: real switches, not just complaints; workarounds people built themselves; a struggle you can predict before they describe it. Past Emerging, PULL has done its job, and the work shifts to sizing, scoring, and segmenting.
06 · The method
Five steps from an idea to what to build
Immerse
Get close enough to the job to feel the struggle the way the people doing it feel it. Immersion works two ways, and you need both.
Observation shows you the struggle. The interviews show you whether it repeats.
I want to feel the pain of the person the same way they feel the pain.Linear's head of product, as quoted by Lenny Rachitsky
Formulate
Write a PULL hypothesis for one specific person, from what you observed and heard, not from what you assumed. Fill in all four conditions. If you can't, you don't have a hypothesis yet.
Validate
Test the hypothesis in conversation, first with the person you wrote it for, then with more people, and more companies, that fit the same profile. One conversation is an anecdote.
Snyder's two-part conversation: pull discovery first, then a small pitch. Open with "Why did you take this call?" Sometimes they describe the whole PULL hypothesis on their own.
Reading the responseLeaning in vs leaning out
How to tell pull from polite enthusiasm, and what a lean-out can mean
Snyder separates two kinds of enthusiasm. Pull leans toward a decision: they ask how it works, how to get started, and what it costs. Leaned-out enthusiasm sounds just as friendly but pushes the decision away, with "tell me more" or "send me more information."
- "Wait, you can do that?"
- "How do we get started?"
- "What does it cost?"
- "That's the future of our industry."
- "Send me more information."
- "Maybe next quarter."
After about five conversations with one kind of person, stop and count the lean-ins, not the compliments. If no one leans in, revise the hypothesis or ask the Hail Mary: "Is there anything on your to-do list you would pay me to do right now?" If one does, run five more with the same profile. When pull repeats, score it.
Score the Struggle
Repeated pull tells you the four conditions are present. Scoring tells you how strongly, and on what evidence. Unavoidable is a yes or no the buyer tells you. Scoring grades it, names the stake behind it, and checks it against what they had already done. The Impact of Delay turns Unavoidable into a stated consequence and a deadline. Lean-ins from your conversations don't count here: score only what people had already done before you arrived. It also makes hypotheses comparable, so you build for the strongest struggle, not the one that feels most exciting.
Pick one, and add Identity when someone's standing rides on it. This defines why it's unavoidable: what motivates them to act and the outcome they expect.
Three or four checked is high. The last two open the Impact of Delay: what happens if they wait, and when it must be addressed.
Pick the highest thing they had already done before you arrived.
- 3Strong signal. High urgency, strong evidence. The stakes and the proof together.
- 2Early signal. High urgency, weak evidence. Severe, but caught before people have acted.
- 1Mixed signal. Low urgency, strong evidence. Real but survivable: people live with their workaround.
- 0Weak signal. Low urgency, weak evidence. Unlikely to create demand.
Need the full card? Score one struggle with the interactive rubric.
Build, but don't stop there
When PULL holds, buyers don't stick with the status quo, and they don't wait to buy: the project is unavoidable and the options they looked into fall short. That settles the first two of your four customer decisions. The other two happen after you build, and they decide whether the growth you won stays.
PULL confirms they act. The project is unavoidable, so standing still isn't an option.
PULL confirms they buy, and soon. The alternatives fall short, and it can't wait.
Yours to earn after the sale: get every customer to the moment Snyder calls "weird to churn."
Yours to earn before the first renewal: the progress has to hold up enough to keep paying.
Snyder is right that pull is found, not manufactured. The question before you build is how strong it is, how proven, and where it breaks.
Questions founders ask
PULL, The Power of Pull, and finding demand before you build
What is Rob Snyder's PULL framework?
PULL is a test for whether a specific buyer will pull your product out of your hands. Rob Snyder, author of The Power of Pull, says a buyer has pull when four things are true at once: there is a Project on their to-do list, it is Unavoidable right now, the options they Look into are Lacking for some critical reason. When all four hold, it would be weird for them not to buy.
Is Rob Snyder's pull the same as the Pull force in Jobs to Be Done?
No. Snyder's pull belongs to the buyer: a project they must get done now, with no option good enough, and it exists before any product shows up. In JTBD switching behavior, the Pull force belongs to a solution: the attraction of a specific new solution, weighed against the buyer's Anxieties about it. John Gusiff captures the buyer's full state in a Desire-for-Progress Statement, a six-forces sentence in which Snyder's Project and Unavoidable map to the Trigger and Push, and the Pull force is its own slot.
What does PULL stand for in The Power of Pull?
Project, Unavoidable, Look, and Lacking. On his slides, Snyder writes the two Ls as the options a buyer Looks into and why they are Lacking. In conversation, including on Lenny's Podcast, they are often described as the List of options and their Limitations. Both mean the same thing: the buyer has looked at alternatives and none of them gets the job done.
How do I know if customers will buy my product before I build it?
Look for behavior, not agreement. Play the struggle back to people with Snyder's two-part conversation and watch for buying signals like "How do we get started?" and "What does it cost?" When pull repeats across people, score the struggle: what is at stake, how urgent it is, and what people had already done about it, such as searching for alternatives, building a workaround, or paying for something.
What is the difference between pull and customer enthusiasm?
Direction. Pull leans toward a decision now: buyers ask how to get started and what it costs. Enthusiasm that leans out sounds friendly but pushes the decision away, with replies like "send me more information" or "maybe next quarter." Lenny Rachitsky calls that kind of response a polite exit.
How many customer conversations do I need to find product-market fit?
Treat five as a checkpoint, not a sample size. After about five conversations with one kind of person, stop and review. If none show pull, revise the hypothesis. If one does, run five more with the same profile to see whether it repeats. Repeated pull is the signal worth building on.
Why don't design partners turn into paying customers?
Design partners often show plenty of interest with no evidence behind it. Snyder notes they tend to give feedback without ever buying, because people with an urgent problem want it solved, not a seat on your product team. On the Score the Struggle rubric, that is high stated interest with no workaround, spend, or switch.
How does the PULL framework relate to Jobs to Be Done?
Closely. Snyder describes a Project as a job to be done, and each PULL condition surfaces at a stage of the JTBD switching timeline: the struggling moment, the trigger that makes it unavoidable, active looking, and deciding. Customer Progress and No-Decision interviews reconstruct those stages from people who have already acted, or who stayed put.
How is Score the Struggle different from PULL?
They answer different questions. PULL helps you recognize a buyer who needs a better option now. Score the Struggle measures how strong that need is, by scoring the stake, the urgency, and the evidence of what people had already done before you arrived. Validate with PULL conversations first, then score the struggles that show repeated pull before you build.
What if prospects agree the problem is real but don't buy?
There are two common causes. If nothing makes it urgent, the struggle is real but tolerated, which is Latent demand: look for the trigger that would make it unavoidable. If the urgency is there but the buyer hesitates, it is often anxiety or an unaligned buying group: ask what would need to be true to move now. No-Decision interviews separate the two.
Can someone run these customer interviews for us?
Yes. Customer Centric Solutions LLC runs Customer Progress, Switching, and No-Decision interviews for Seed through Series B founders. Buyers are often more candid with a neutral third party than with the founder selling to them. Book a discovery call to scope it.