How Continuous Discovery Works (and Doesn’t) in Early-Stage Startups

Teresa Torres argues that continuous discovery habits apply to early-stage startups, not just mature products, even before a product or customers exist. She illustrates the ideal path with fictional founders who set a directional outcome, interview target customers, and run assumption tests. She then explains why many startups start from an idea instead, which fuels confirmation bias and escalation of commitment. Her practical advice covers finding pre-customer interviewees by going where they gather and paying to reach them. The core point is that outcomes and discovery should follow the value proposition and customer, not the idea.

01Key takeaways

• Set a directional outcome that bounds discovery even when you cannot yet measure success.
• Start from a customer segment and value proposition rather than a fixed product idea.
• Interview both novices and experienced customers to map the opportunity space quickly.
• Watch for escalation of commitment and confirmation bias when a team is invested in one idea.
• Story-map an existing solution to generate assumptions, then test the riskiest ones first.
• Find pre-customer interviewees by joining communities, observing, and giving help before asking for time.

02Key sections

The ideal discovery path

Two fictional founders pick a customer segment and value proposition, then use interviews and assumption tests to explore the opportunity space. The example shows discovery working before any product exists.

Directional outcomes over vanity goals

Teams can set a non-measurable directional outcome that scopes discovery and evolve it into a measurable one later. Goals like 'find customers' or 'reach product-market fit' do not bound the opportunity space.

Why startups start from ideas

Founders who begin with a solution tend to frame discovery as validating the idea, which leads to escalation of commitment and confirmation bias. This makes negative evidence easy to miss.

Breaking the cycle inside idea-led startups

Product people can map the opportunity space by interviewing, then story-map existing solutions to surface and test risky assumptions. Findings are brought back to founders to evolve the solution.

Recruiting customers before you have a product

Observe and engage in communities where target customers already gather, then pay ads to drive traffic to landing pages for interviews and tests. B2B teams should adapt channels to associations and conferences.

03From the post

““I get that the continuous discovery habits framework works well for mature products, but does it work for early-stage startups?” This question always surprises me. I spent all of my full-time employee experience at early-stage startups (many of them pre-product) and I relied on these same”

04Frameworks mentioned

Summary and takeaways written by PM Atlas; quotes are short excerpts. © the original author.