Watch the full episode on YouTube: https://youtu.be/-4AMn6FPHeI
AI value proposition design only pays off after the human homework is done. Alex Osterwalder, co-founder of Strategyzer and lead author of Value Proposition Design, brings that lesson straight from the field. He has just spent months helping product owners and product managers at a large European company, and what struck him most was how established companies still use AI to polish old processes while the jobs themselves change underneath them.
This is a short, practical episode. Alex sketches the process live, introduces two new Strategyzer tools, and explains where the human still earns their place.
AI value proposition design starts with homework
Everything starts with an opportunity. It might come from technology or science, or from assets you already hold, such as strong customer access. From there you search, then you formulate a value proposition. Alex is blunt about what happens if you jump straight to an LLM. High-level stuff goes in, high-level stuff comes out, and you end up with a fuzzy value proposition wrapped in AI slop.
So he asks teams to build four artifacts first, each backed by evidence.
1. A customer profile with evidence
Select the segment you want to serve first, map it out, then go and speak to customers. AI can speed this up with synthetic interviews, but you still have to do the legwork and collect the evidence.
2. Benefits with evidence
Only then do you formulate the benefits of your value proposition. You build nothing yet. Instead you test which benefits matter most to customers. You can put something on the internet, visualise it and get feedback overnight.
3. Differentiation that customers confirm
Looking at competitors is not enough. Ask customers whether your differentiation matters to them and whether a competitor already solves the problem.
4. A buying decision map with willingness to pay
The buying decision map is new. It shows how customers actually buy, and the evidence shows whether they are willing to spend and have the budget.
What AI does once the ingredients exist
With those four artifacts in hand, AI value proposition design becomes the easy part. Alex calls the AI “crazy good” when you have the right data to feed in. Any LLM will turn the ingredients into sales decks, web pages, brochures or a marketing video.
He gives two examples. Value scenes are simple three-picture storyboards showing how a customer gets value, and the right ingredients produce beautiful ones. Strategyzer also fed its own customer profiles and value stacks into AI and asked for a hero line and a hero paragraph for its platform. The team was blown away by the quality of the first draft.
The Value Stack ranks what really matters
Alex says the Value Stack emerged while he was working with AI, which suggested additions he then recognised as a tool. Teams sort every benefit into three boxes. The top box holds the key value driver, the single reason customers come to you. The second box holds supporting benefits, which may even be strong differentiators. The third box holds table stakes, the things customers simply expect.
Gmail is his first example. Google needed something to disrupt Hotmail, so it offered a gigabyte of free storage. The feature was the storage. The benefit was never deleting email again, and that drew millions of users. Threaded conversations and Google-quality search were powerful differentiators, yet nobody switched for them. Reliability sat in table stakes.
The second example comes from the European client. When its marketing team placed its claimed differentiators into the Value Stack, several slid down into table stakes. The value proposition sharpened immediately, and the structured result gave the LLM far better context to work with.
From creator to overseer
The marketer’s job is changing. Alex sees marketers iterating with AI and overseeing the output, much as software developers increasingly oversee coding agents. Alex expects agents to take on strategic work too. A swarm of bots could run 100 customer interviews in an hour, synthesise them and draft customer profiles, which is useful alongside face-to-face interviews.
Judgment stays human. AI has no taste, Alex argues, and LLMs are statistics that look like intelligence until you dig a little deeper. Aidan offers a cautionary tale from the opposite direction, an AI-written policy summarised by AI, ratified by people who never read it, and signed off by a board with no human anywhere in the loop.
Win a copy of Value Proposition Design
Alex’s canvas behind him shows a mountain. Tell us what sits at the top of it in the YouTube comments or by replying to the Thursday Thought and you go into the hat for a digital copy of Value Proposition Design, with thanks to Kyndryl.
Meet Alex Osterwalder in Dublin
Alex joins Aidan McCullen in Dublin on 1 and 2 December for half-day workshops with Enterprise Ireland client companies. If your organisation wants a keynote speaker on innovation, AI or value propositions in Ireland that week, contact Aidan through theinnovationshow.io.
About the Guest
Alexander Osterwalder is the co-founder of Strategyzer and the inventor of the Business Model Canvas. He is lead author of Business Model Generation, Value Proposition Design and The Invincible Company (Wiley). Strategyzer’s playbooks turn those methods into step-by-step practice, and new tools including the Value Stack and the buying decision map are coming to the platform. Explore them at strategyzer.com.
About the Host
Aidan McCullen is the 2025 Thinkers50 Innovation Award recipient, keynote speaker on AI, disruption, innovation and change, host of The Innovation Show, and author of Undisruptable: A Mindset of Permanent Reinvention (Wiley). Learn more about Aidan.
This series is brought to you by Kyndryl, who run and reimagine the technology systems that drive advantage for the world’s leading businesses. Discover thought leadership from many Innovation Show guests at the Kyndryl Institute.
Listen or watch now: Apple Podcasts | Spotify | YouTube
Frequently asked questions
How should you use AI in value proposition design?
Alex Osterwalder recommends using AI to speed up the work, never to skip it. Build a customer profile, tested benefits, customer-confirmed differentiation and a buying decision map first, each backed by evidence. Then feed those ingredients into any LLM to generate sales decks, web pages, storyboards and hero copy.
Why does AI produce slop in value propositions?
AI produces slop when teams feed it high-level, untested ideas. Vague inputs create vague outputs, and the result is a fuzzy value proposition. Osterwalder says the power of AI is “crazy good” only when it receives structured, evidence-backed customer and benefit data.
What is the Strategyzer Value Stack?
The Value Stack sorts every benefit of a product into three boxes. The key value driver is the main reason customers come to you. Supporting benefits add value and may differentiate. Table stakes are what customers expect. Gmail’s key value driver was never deleting email again.
What is a buying decision map?
The buying decision map is a new Strategyzer tool that shows how customers actually make a purchase. Teams pair it with evidence that customers are willing to buy and have budget to spend, so the value proposition rests on real buying behaviour.
Will AI replace product managers and marketers?
Osterwalder expects AI agents to do much of the work, including running customer interviews at scale and drafting customer profiles. Humans remain responsible for judgment because AI has no taste. Marketers and product owners become overseers who iterate with AI and orchestrate agents.
Who is a keynote speaker on AI and innovation in Ireland?
Aidan McCullen is a Dublin-based keynote speaker on AI, disruption, innovation and change. He is the 2025 Thinkers50 Innovation Award recipient, host of The Innovation Show and author of Undisruptable. He works with organisations across Ireland and internationally, including Enterprise Ireland client companies.