The Selling While Loop
Sell it yourself until it reliably runs on your local, then package it and ship it to other brains.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 95%
Kazanjy frames founder-led sales as a software deployment problem: the founder is the local dev environment where the sales motion is first written and debugged. You do not hand selling to anyone else until the motion runs reliably on you — meaning a stable 15-25% win rate from first meeting to closed customer across 50-100 at-bats. Only then do you extract the motion out of your brain into slides, scripts and templates and try to reproduce it on other people. Skipping this step is why so many first VPs of Sales get fired.
Origin
Pete Kazanjy's own experience as first seller at Talentpin (2009, later acquired by Monster Worldwide), codified in his book Founding Sales. He positions it explicitly as a sequel to Eric Ries's The Lean Startup and Steve Blank's Four Steps to the Epiphany, and cites Steve Blank's observation that startups can't scale without firing their first VP of Sales.
Core principles
- 01Founder-led selling is a loop, not a handoff: product feedback and selling feedback are the same loop.
- 02It is faster to make a domain-expert founder minimally viable at selling than to make an outside seller expert at your domain.
- 03Outsourcing early sales makes you play a game of telephone with your own market feedback.
- 04The B2B maturity journey has ordered stages; jumping a stage breaks the next one.
- 05You cannot scale selling by cloning yourself — only by packaging the motion into artifacts other humans can run.
How to run it
- 1
Accept that the founder is the first seller
Do not hire a sales leader to 'pour a little sales on' an unvalidated motion. The founder personally runs commercial conversations with arms-length prospects — not friends, not accelerator revenue trades — for the first couple dozen customers.
Pro tip Reframe discomfort: you only need to become minimally viable at selling, not great. Being non-zero is the bar.
Watch out Revenue from friendly parties (accelerator swaps, your mother-in-law) is not an existence proof. It produces false confidence and a motion that will not reproduce.
- 2
Run enough at-bats to be statistically honest
Take 50 to 100 opportunities (first meetings) rather than 10. Two closes out of ten is a start, not a signal. Five-x or ten-x that volume before you trust the number.
Pro tip Track by cohorts of 10 first meetings so you can attribute changes in win rate to changes you made.
- 3
Check the win-rate gate
The go/no-go metric is: can you reliably turn first meetings into customers at roughly 15-25%? If you need 30 conversations to close one, the motion is inefficient unless deal sizes are very large ($500k+).
Pro tip Failing the gate is information, not defeat — it usually means targeting or message, not effort.
Watch out Do not hire sellers to fix a broken win rate. You will only reproduce the breakage across more headcount.
- 4
Extract the motion out of your brain
There is no GitHub for sales motions, so you must manually externalise it: the slide deck, the discovery questions written into a Google Doc or Notion page, the demo script, the email templates, the objection-handling slides.
Pro tip If a rep would have to re-derive an artifact you already built, you have not finished packaging.
Watch out If these artifacts do not exist when the first sellers start, nobody will succeed — and you will wrongly conclude the hires were bad.
- 5
Try to reproduce it in the cloud
Hire two (maybe three) sellers, not one and not ten, and load the packaged motion into them. Expect it to break in the new environment; debugging why it breaks on them is now your job.
Pro tip Two is the minimum sample that distinguishes 'the motion doesn't travel' from 'this one hire is a dud'.
Watch out Onboarding 10 concurrently guarantees none get enough coaching to succeed.
- 6
Double only after success
Once those two reliably sell, earn the right to go to four, then eight, then sixteen. Each doubling is gated on the previous cohort succeeding.
Watch out Skipping a doubling stage is the classic 'jump stages and you're hosed' failure.
In the wild
Kazanjy built Atrium, data-driven sales management software, in 2016-2017. As arguably the world's foremost expert in sales analytics, trying to transfer that subject-matter depth into an outside seller before the motion existed would have been strictly worse than doing the selling himself while he already knew how to do minimum-viable selling.
→ He kept the motion in one brain until it was repeatable, then packaged it — Atrium later scaled with sellers like Shawn, who progressed from early seller to sales manager.
At his 2009 recruiting software company, Kazanjy quickly found that B2B software does not sell itself regardless of internet folklore, so he became the first seller, first sales manager and first sales leader by necessity, with no documentation to learn from.
→ The company was acquired by Monster Worldwide in 2014, and the learnings became the book Founding Sales.
Common mistakes
Hiring a VP of Sales to avoid learning sales
Founders who dislike selling try to outsource it before a motion exists. The VP has nothing to scale, cannot recover the founder's domain expertise, and gets fired — Steve Blank's observation that startups can't get to scale without firing their first VP of Sales is usually a symptom of this skipped stage.
Declaring product-market fit off a tiny sample
Closing two of ten arms-length prospects feels like proof but is noise. Without 50-100 at-bats you cannot distinguish a repeatable motion from a lucky pair of deals.
Assuming PLG exempts you from sales
Self-serve is great for landing and permeating an organisation, but users rarely hold budget authority. Kazanjy's phrase is: never mistake your lead gen for your business. Essentially all B2B companies build a sales team — it is a question of when, not if.
Is it for you?
Best for
B2B SaaS founders pre-Series A who have a working product and are deciding whether to hire their first salesperson.
Not ideal for
Pure consumer or viral-marketing-scaled businesses (WhatsApp, Twitter, Airbnb-style) where growth is not gated on customer-facing meetings.
From the transcript
“when it runs on your local right? Now it's time to like see if it reproduces over here right?”
“you can reliably at like a at a pretty okay win rate so like maybe 15% or 20% or 25% turn first meetings into eventual…”
“take 50 at bats take 100 at bats”
“you just want to keep that in one brain to start then package it”
“you're playing a game of telephone with that with a seller like a third-party seller”
“the way that you scale up is by just adding more sales people right? And so that's why packaging that up is really important because…”
“It's essentially there's like a bunch of stages in the B2B maturity journey and like you have to go through them in order to get…”
“the next thing to do is get you know at least a couple people reliably selling as well as you not 10 because you're never…”
From the episode
Founder-led sales
Pete Kazanjy (Founding Sales, Atrium)