Productize to Segments (the 'You Can Act Differently' Test)
Segments are defined by needs and willingness to pay — and only count if you can act differently on them.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 93%
About 60% of companies claim a segmentation strategy; roughly 10% actually have one. Most do demographics or personas, which do not predict what someone will buy. Ramanujam's test: a segment is only real if you can act differently on it — build a different product, a different business case, a different message, a different price. Segmentation must be grounded in what customers need, what they value and what they will pay for, and the output is productizing to segments, never positioning one product at several segments.
Origin
Madhavan Ramanujam's reframing of classical segmentation, developed at Simon-Kucher and set out in Monetizing Innovation. His illustration of why persona-based segmentation fails: a 70+ year old, incredibly wealthy man living in a castle in the UK describes both King Charles and Ozzy Osbourne — same demographic, entirely different needs and willingness to pay.
Core principles
- 01Segments are need-based and WTP-based, not demographic or persona-based.
- 02The test of a real segment: you (product, sales, marketing, finance) can act differently on it.
- 03No market has homogeneous needs — heterogeneity exists whether you accept it or not.
- 04One-size-fits-all is one-size-fits-none.
- 05Segmentation is done up front; launching for every segment is not — pick one segment to serve first.
How to run it
- 1
Map heterogeneity of need, not of demographics
In the willingness-to-pay conversations, ask who is willing to pay for this innovation, what they need, what they value, and how many of them there are. Cluster on those answers, not on age, industry, or title.
Pro tip The water test: fountain water is free, a bottle is $2, sparkling is $2.50, the minibar is $5. Same water — different needs, different packaging, different price.
Watch out A persona is not a segment. Charles and Ozzy Osbourne are the same persona and have nothing in common as buyers.
- 2
Apply the act-differently test to every candidate segment
For each proposed segment ask: would we build a different product, a different business case, a different marketing message, a different sales strategy, a different price? If the answer is no across the board, it is not a segment.
Watch out If your reaction and treatment are identical across 'segments', you have labels, not segmentation.
- 3
Size each segment and check it can support a product
Establish that each need-based cluster represents a meaningful total addressable market with similar needs and similar willingness to pay before committing resources to it.
- 4
Sequence: pick one segment, productize for it, then add the next
Early-stage companies cannot ship five products for five segments. Use the segmentation to prioritize R&D and resourcing — which segment first, which product first, which value message first — and add the others later.
Pro tip Doing the exercise early tells you how many segments exist, their size, and the order to attack them. That is the real payoff, not shipping a full product line on day one.
Watch out The lazy path is to build everything into one product and try to attract everyone. You will attract no one.
- 5
Graduate to dynamic segmentation
The same person belongs to different segments at different moments — Uber Comfort with 'quiet preferred' when working in the car, Pool for a routine commute; pizza on Friday night, a healthy choice on Tuesday lunch. Model the situation, not just the person, and productize to the moment.
Pro tip This is the frontier: get static segmentation right first, then layer a dynamic view.
In the wild
Apple did not build one iPhone at one price to maximize market share. There is an iPhone at 299, 399, 499, up to 1,499. Ramanujam himself walked into the store for an iPhone X, didn't want the retina features, saw an 8s at 799 and walked out with that instead.
→ Apple productizes to different needs segments rather than pricing one product, and is the most profitable company on the planet.
Pre-IPO, Eventbrite served all customers with a single product. Simon-Kucher ran a segmentation on needs and WTP: a local wine-club meetup only needs one general-admission ticket type; a professional event organizer needs unlimited ticket types (GA, VIP, etc.).
→ They productized into essentials / professional / enterprise plans, with the single-ticket-type product priced cheaper — capturing both segments instead of compromising for both.
Common mistakes
Persona-based segmentation
Segmenting on demographics or personas tells you nothing about what someone needs or will pay for, and produces messaging exercises rather than products.
Building one product and positioning it at multiple segments
If you build one thing and try to message it four ways, you have already lost the battle — none of the four groups gets what it actually needs.
Deferring segmentation because you're early and busy
'We have no time and no resources for multiple products' is a misunderstanding. Segmentation done early is what tells you which single product to build first and for whom.
Is it for you?
Best for
Product and growth leaders about to build a product line, add plans, or expand beyond a single one-size-fits-all offering.
Not ideal for
A single-feature product with a genuinely narrow, homogeneous buyer where added SKUs would only create complexity.
From the transcript
“segmentation needs to be based on what customers need what they value and what are they willing to pay for”
“walk into these companies they'll say we are building a one-size-fits-all I would quickly collect them and say one size fits none”
“it's the same damn water but it's packaged productized differently because people have different needs”
“the three most important words and what you said is you act differently”
From the episode
The art and science of pricing
Madhavan Ramanujam (Monetizing Innovation, Simon-Kucher)