Learn, Don't Earn: The Zero-to-One Product Rule
- Difficulty
- Easy
- Time to result
- ~months to results
- Steps
- 7
- Confidence
- —
Partaker's decision rule for the moment a business wants to replace a working revenue stream with an unproven one. The rule has two halves. First, never kill what earns while you are still learning — the existing product keeps funding the business until the new one is proven. Second, and less obvious, change the objective for the new product from 'make it as good as possible' to 'learn as much as possible about what it should be'. In the zero-to-one strategy formulation stage, strategy is less about getting it right beforehand and more about organically discovering it once you get moving. He is explicit that this does not mean asking customers what they want — the Ford faster-horse and Steve Jobs iPhone objections still hold — but that your own visionary ideas will be organically informed by what you learn from the assembled audience.
Origin
Extracted from Deep Dive with Ali Abdaal
How to run it
- 1
Separate the earner from the learner
Identify which product currently funds the business and which is the unproven bet. Name them explicitly: don't kill earn while we're trying to learn.
Watch out Killing your main revenue stream before the next product is proven is, in Partaker's words, absolutely nuts.
- 2
Freeze improvement work on the earner
Stop investing team time in making the earning product better while its future is undecided. If people are getting results, giving good reviews and rarely asking for refunds, it is good enough for now.
Pro tip Redirect the freed capacity straight into the learning product rather than letting it disperse.
- 3
Restate the new product's objective as learning
Replace 'make it as good as it can possibly be' with 'learn as much as possible about what this should be'. The offer, the pain points and the interests are all unknown until the group assembles.
- 4
Assemble the audience first
Get the people who will buy into a room with initial glue around the product, then spend time with them to find the natural segments, their pain points and what they actually want to learn.
Pro tip Segments discovered this way inform the product suite and potentially separate value ladders.
- 5
Buy the right to change it
Structure the offer so you retain freedom to change. Sell a short commitment rather than a long one, or tell buyers openly that this is a beta, things will change, and a refund is available at any point if they want one.
Pro tip Transparency at signup is what converts customer goodwill into permission to iterate.
Watch out A twelve-month program taken at full price makes founders feel unable to remove features, because customers have paid and will be justifiably annoyed.
- 6
Reserve the call option on the earner
Make no irreversible decision about the existing product. Keep servicing existing clients through their term, and preserve the option to relaunch it unchanged if the new product underperforms.
- 7
Decide the merge only after the data arrives
Once the new product has real buyers, use what you learn to decide whether and how far to merge the old offering into it — including whether the audience circles overlap enough to bundle at all.
In the wild
Ali sold roughly 250 people into a $5,000 twelve-month accelerator based on the team's best guess at what would be valuable. Fairly quickly they saw that students used features A and B far more than C and D, while C and D consumed most of the team's time. But because customers had bought a twelve-month package and the company's core value was to put students first, the team felt unable to remove anything mid-term. The long commitment, not the wrong guess, was what blocked iteration.
→ Ali proposed two structural fixes for the next product: a one-month offering, or an explicit beta framing with a refund available at any point.
Partaker applied the same rule to his own business rather than only prescribing it. With nearly half a million followers on LinkedIn and a newsletter around 145,000, he had a newly started chief product officer whose brief was to discover what the offer should be by spending a lot of time with that audience — finding the natural segments, their pain points and what they wanted to learn.
→ The segments discovered would then inform the product suite and potentially different value ladders, rather than the ladders being designed up front.
Common mistakes
Killing the goose before the replacement sells
Ali's productivity product had zero revenue and had not launched, while the YouTube ladder was funding the business. Partaker's response was blunt: it is nuts to think of killing your main revenue stream without the next product proven. Necessity-is-the-mother-of-invention arguments do not justify it.
Optimising a product you can't yet define
Pouring effort into making the new thing as good as possible presumes you know what it should be. In the zero-to-one stage you do not, and the polish is spent on guesses that the first real cohort will overturn.
Reading 'learn' as 'ask them what they want'
Partaker pre-empted the Ford and Jobs objection directly. Learning mode does not mean the customers dictate the roadmap — you may still produce your own Model T or iPhone idea, but it will be organically informed by what you picked up along the way.
From the transcript
“the go is to learn not to earn”
“I think it's just absolutely nuts to think of killing your main Revenue stream right from a product point of view without like the next…”
“because you're in the zero to one you know strategy formulation stage where strategy is figured out it's less about trying to get it right…”
From the episode
A Brutally Honest Conversation With My CEO Coach - Eric Partaker
Eric Partaker