Teams inside established companies
The intrapreneur’s three advantages, from my 2018 talk
My talk “How to be an intra-preneur” argued that a team building something new inside an established company starts with three things: real paying customers, a business model it understands, and a network of untapped resources. I think that talk is even more salient now, so here it runs slide by slide, updated for 2026.
From a talk I first gave in 2018, updated for 2026.
Advantage 1: existing paying customers
Slide 18 put it in four words: get to know them. Slide 19 named three ways in: personas and psychographics, existing customer feedback, and jobs to be done.
Slide 28 asked the questions a team inside a company can answer this week: “Does your company track this stuff? Do you have a customer service desk? What are the most common complaints?” Then go ask yourself, your team and your early adopters what you do well, what you do poorly, and what you could change.
The 2026 update: the existing customer data, customer touch points and customer insights can now be mined on a continuous basis. Artificial intelligence (AI) reads support tickets, survey comments, sales call notes, and renewal and churn reasons in the same week, inside your company’s data policy, and codes them into what pushes customers to change and what holds them back.
Slide 41 gave the format for writing down what you find. As a (persona), when (urgent event) happens, I would like to (solution), so that I can (solve the job to be done).
Advantage 2: known value and growth hypotheses
Slide 44 put Eric Ries and Steve Blank side by side. Value hypothesis = product-market fit. Growth hypothesis = channel-market fit. Slide 45 laid out Blank’s four steps: customer discovery for problem-solution fit, customer validation for product-market fit, customer creation for channel-market fit, then company building to scale.
Slide 46 set the bar: different enough, that matters enough, that someone will pay you. Slide 50 asked what every new line has to settle first: “What’s your riskiest assumption??”
Slides 52 and 53 showed the answer in practice, with the Medlantis experiment board from 2016: about nine months of tests, each with a pass mark set before it ran. One row came straight from our customers’ systems. Sales trainers needed to track completion in their learning management system (LMS), so the next test asked whether we could build compliance with the Sharable Content Object Reference Model (SCORM), the tracking standard those systems use. We could, so we persevered. The board, row by row, on the method page.
A team inside a company knows how its business makes money, so it can write both hypotheses for the new line in plain words on day one, then score each assumption on the chance it’s wrong times the damage if it is.
Advantage 3: an untapped internal network
Slide 59 named the content trap: “It’s about what people do with what you build.” Slide 60 turned to network effects, and slides 62 to 64 to complements, with the question I’d still put to any team building a new line:
“Where along the user journey - the Jobs to be Done - are your customers employing substitutes that could easily be fulfilled by a new feature, product, or service that you could build (or buy)?”
Slide 64 laid out three phases: stand-alone value, then networked value, then scale. For a team inside a company the network starts close to home: the sales team and account managers, partners and distributors, the events the company already exhibits at, its newsletter and content audience, its experts, and the complements its customers already buy. Channel tests start from those, then move to strangers.
The steps every internal team works through
Each of the three gives the testing a place to start. None of them proves demand for the new line on its own, so each becomes a row in the assumption portfolio: “customers who buy our current line will buy the new one”, “account managers will introduce it”.
The situation brings its own habits and anxieties, and the team, the sponsor and the sales team share them. The budget review arrives on a date the planning calendar set. The accounts the new line needs belong to a sales team with a quota for the current products. Anything public carries the company’s name, so tests go through the brand team. The customer’s systems decide how a new product fits: at Medlantis we had to build our product to work with each customer’s training systems and the SCORM tracking standard before it got out the door. And buying from an outside supplier runs through procurement, legal and information technology (IT) security.
That situation sets the order of the work: existing accounts first with the account owner in the loop, the company’s approvals built into the plan, and evidence ready for each review.
The sources behind the talk
The talk drew on published work: Steve Blank’s customer development, from The Four Steps to the Epiphany (2005); Eric Ries’s value and growth hypotheses, from The Lean Startup (2011); and jobs to be done, which traces to the 2007 paper by Clayton Christensen, Scott Anthony, Gerald Berstell and Denise Nitterhouse, “Finding the Right Job for Your Product”. For the framework nerds, the full lineage runs back to Kurt Lewin in 1936.
Starting points: the scorecard and the kit
Start with where your next five customers will come from: 13 multiple-choice questions, about 2 minutes. If the new line has no paying customer yet, the free kit runs this method in your own AI assistant; use the one your company approves for customer data. If it has its first customers, see how the work runs inside a company, or tell me your origin story.