Inside the room
Stop Convincing, Start Finding: Rob Snyder on Pull
Startup advice as bloodletting
Rob opened with bloodletting. It was a real medical practice, built on years of learning and subtle dogma, all resting on a wrong theory of what causes illness. It made patients worse, and it seemed right at the time.
His claim is that a lot of startup advice is the modern version: careful practices built on the wrong theory of what causes growth. His own list of things that seemed important and weren’t includes startup events and pitch competitions, accelerators, VC feedback, mission statements, business model canvases, “validation” exercises, and a lot of discovery interviews.
Pull comes before the product
Rob defines pull as a customer ripping your product, or even your product idea, out of your hands. Its opposite is push: convincing customers to buy and use something, which he likens to pushing a boulder up a hill.
What causes pull is priorities, not pain. Nobody drops their priorities because you described their pain points well. Pull is a priority they are trying to get done right now that their existing options can’t handle. Rob turns it into a sentence to fill in: “I am trying to do X right now because Y, but can’t do X with Z because A.” Each letter of PULL is one blank:
01
Project.
Something on the customer’s mental to-do list.
02
Unavoidable.
It has to happen right now.
03
List of options.
The ways they could get it done today.
04
Limitations.
Why those options don’t get it done.
Rob calls the result a state of “unacceptable stuckness”: doing nothing isn’t OK, and neither are the existing options. It exists before the customer has heard of you, so you find it rather than create it. When it’s there, buying your product is close to mandatory. When it isn’t, no pitch or messaging will create it.
His example was Jump, an AI notetaker. Financial advisors have to take meeting notes, the generic tools didn’t connect to their CRM, and skipping notes is an SEC compliance risk. When Jump showed up, advisors grabbed it. Insurance agents looked the same on paper, but nothing happens to them if they skip the notes, so they simply stopped taking any. Same product, no pull, no sales.
Pull also writes your positioning. Rob’s template: “We offer {product category} that enables you to overcome A to accomplish X.” For Jump, that meant an AI notetaker for financial advisors that integrates with their CRM and won’t get them thrown in jail.
Pain points and ROI don’t cause pull
Rob’s list of things that seem like they should cause pull, and don’t, includes a hair-on-fire pain point, jaw-dropping ROI, a new category, happy customers, and a product 10x better than anything on the market.
One example: European supplement companies get fined constantly by the EU over their ads, and call it the most annoying thing ever. A $100-a-month AI tool that vets ads for compliance still didn’t sell. The fines were already in the marketing budget, and they had other priorities. As Rob quotes Bob Moesta, “Bitchin’ ain’t switchin’.”
Value doesn’t do it either. Damballa pitched a cybersecurity tool to eBay’s CISO, who said it would save them $40M, and still didn’t buy. It wasn’t a priority.
Big pain and big ROI leave buying optional. Pull is what makes it mandatory. For later-stage companies where some customers buy fast and others stall, Rob’s diagnosis is the same: some have pull and some don’t, so stop selling to the ones who don’t.
Finding pull at the idea stage
Rob’s advice is to get first-hand experience: job shadowing, consulting, doing the work alongside the people you want to serve. His slide on discovery interviews calls them “an offense against God.” Design partners don’t fare much better. He describes the arrangement as asking someone to spend an hour every week, indefinitely, grading your homework.
One shortcut is a Hail Mary question: is there anything you’d pay someone right now to solve? One cold call built on that question turned a Disney executive into a first customer, closed in three weeks.
The only test of whether pull is real is whether someone tries to buy, even before the product exists. “This would be so valuable for our team!” is not pull. “I need this right now, how do we get started?” is.
His example was a YC startup that, after twelve months of hell, pivoted from devtools into insurance with no idea what to build. They emailed insurance brokers offering to fly out and build whatever they wanted for $500, did it with three brokers over three weeks, and found a common need: generating proposals for prospective customers in minutes rather than days. Cold calls to other brokers then closed three or four out of five.
His rules of thumb for moving fast:
01
Find three to five people with pull, rather than going deep with one and getting pigeonholed.
02
Aim for five conversations in two or three days.
If none of them pull, debug the hypothesis.
03
In legacy industries like manufacturing, operators are often fine with the status quo.
Go to the C-level instead.
04
In an industry you don’t know, shadow people for a week at a time and work down a list.
The pull theory of sales
Most sales advice follows what Rob calls the push theory. The deck explains how you see the problem, why it’s big, why it’s urgent, and why the buyer is in trouble if they don’t solve it. He told the story of a prospect who took a Challenger-style pitch and handed it to a competitor to build instead.
In the pull theory, someone who is stuck just needs something that unblocks them, and will buy as long as you don’t talk them out of it. Educating or convincing them wastes their time, and anything tangential gives them more to object to. Hence what Rob calls the zen of sales: lean back, sit on the same side of the table, and see whether they’re stuck.
How he runs a call:
01
Ask why they decided to take the call.
The answer usually surfaces the pull, and earns you the right to ask questions without it feeling like an interrogation.
02
Fill out the pull framework with them, live.
03
Pitch in two slides: the pull you hear from customers like them, then what you offer and what their life looks like unblocked.
04
Keep the demo optional and short, 30 to 90 seconds, showing their project after they buy rather than how the product works.
Then see if they pull. The path to product-market fit is a loop of about five calls: did they have pull, did they think what you offered was a fit, and did they try to buy?
Outreach follows the same logic. The point of a first message is to seem interesting, not to sell, and to get people to chat with you like a human. Rob shared a short note that got a 30% reply rate, followed by a bump on day three and just their first name on day six.
In enterprise deals, find a champion first: someone who will fight internally to buy. Once they say yes, the rest is project management, a list of steps with dates.
Investors have pull too, and it’s different
Investor pull isn’t “I want to invest in good companies.” It’s “I don’t want to miss the next big thing.”
So Rob suggests opening a pitch with the ambitious version of the vision, then coming back down to reality. Most founders do the opposite and lose the investor before the story takes off.
His rule: keep what investors want and what it takes to build a successful startup in two separate boxes.
Run the company as a case study factory
Rob’s model for scaling treats the startup as a factory that produces successful customers: case studies who bought fast, would be weird to churn, and look like your other case studies.
Most startups load that factory with waste: random discovery questions, “our view of the problem,” category creation, group demos, business cases, irrelevant onboarding and features. Each step adds drop-off. His factory has five stations:
01
Discovery.
Fill out the pull framework.
02
Fit.
The product and their unblocked life, in as few words as possible.
03
Confirm.
What they need to know to bring it into their organization.
04
Project plan.
What has to happen, and by when, to show value.
05
Onboarding.
Get them from “weird to renew” to “weird NOT to.”
From there, attack the factory’s bottleneck. Thinking in these terms lets an engineer debug go-to-market like any other system.
On retention, Rob’s rule is that every customer is guaranteed to churn until one specific thing happens, after which it would be weird if they did. Onboarding should be a bullet train to that moment.
What to try this week
01
Write out your customer’s pull: the project, why it’s unavoidable now, the options they have, and why those fall short.
If you can’t fill in a blank, that’s the hypothesis to test.
02
Test it with five targeted conversations.
If nobody tries to buy within two or three days, debug the customer or the problem before building further.
03
Rewrite your outreach around what’s genuinely interesting to the person reading it.
Cut the AI-generated feel, the generic personalization and the jargon, and lead with something specific and human.
Rob ended on how founders fail at this. They get it, then go straight back to what they were doing before, building product or planning a pitch competition. Your calendar, he says, reveals your theory of what will cause your startup to take off.
So his last question is worth keeping on your desk: what would your calendar look like if your life depended on finding pull this week?
About the speaker
Rob Snyder is a fellow at Harvard Innovation Labs and the author of The Power of Pull.
This recap is based on Rob Snyder’s “What causes fast growth” slides and the night’s talk notes. Customer examples and results are as Rob presented them.
