Price-Elasticity Three-Response Model
When a technology makes something cheaper, work out which of three demand responses your market will take
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 88%
A model for forecasting the impact of a cost-collapsing technology. When you make an activity dramatically cheaper, demand does not simply shrink to match — it responds in one of three ways, and it moves through predictable stages of use. Getting the response right tells you whether an industry contracts, holds, or explodes in value.
Origin
Presented by Benedict Evans as an application of the Jevons paradox / price elasticity of demand to AI, illustrated with the histories of accounting, investment banking (Excel), software libraries, and recorded music revenue.
Core principles
- 01Cheaper does not mean less; it usually means more
- 02The three responses: same output for less money, more output for the same money, or more output for more money because new ROI appeared
- 03Adoption moves in stages: first do the old thing but more, then do genuinely new things, then redefine the question entirely
- 04The biggest value is unlocked by the third stage, which is invisible until someone builds it
How to run it
- 1
Identify what just got cheaper
Pin down the specific activity whose cost collapsed — a week of accounting math, writing boilerplate code, distributing recorded music.
- 2
Choose among the three demand responses
Ask whether the market will do the same for less money, more for the same money, or more for more money because a new return on investment now exists.
Pro tip The 'more for more' case (new ROI) is where headcount and spend actually grow — Excel made banking analysis cheaper and banks hired MORE analysts and did more analysis.
Watch out Assuming 'same output for less money' is the default is the classic error that predicts mass job loss and is usually wrong.
- 3
Locate the adoption stage
First people just do the old thing but more of it on the new platform (print emails, put Flickr on mobile). Then they build things only possible with the new tool. Then they redefine the question — Spotify is not an online music store, it is something else.
Watch out Do not stop your analysis at stage one; 'you have to get past we do the old stuff but more.'
- 4
Hunt for what gets unlocked, not what gets cheaper
Ask what wasn't possible before and now is, rather than only asking how the old activity gets discounted.
In the wild
The joke is that before Excel, junior investment bankers worked brutal hours; you would expect the tool to cut their numbers. Instead Goldman's associates still work long hours doing far more analysis than was ever possible before.
→ Demand took the 'more for more' path — cheaper analysis created more analysis, not fewer analysts.
Global recorded music revenue halved from 2000 to ~2015 (the 'why pay $15 for a CD' phase), then recovered to about 75% of the inflation-adjusted peak via streaming (the 'what if $15/month gives you all music' phase).
→ A completely different question in each half of the curve, ending in a redefined product rather than a cheaper old one.
Common mistakes
Defaulting to the 'same for less' response
Assuming cheaper automatically means the market buys the same amount for less money ignores elasticity and produces the wrong forecast of contraction and job loss.
Stopping analysis at 'the old thing but more'
Most of the value comes from the later stages where new things become possible and the question itself is redefined; analysts who stop at stage one miss Spotify, Uber, and Airbnb entirely.
Is it for you?
Best for
Product strategists and investors forecasting how a market reacts when a technology collapses the cost of a core activity
Not ideal for
Markets with hard non-price constraints (regulation, physical capacity) where demand cannot expand regardless of cost
From the transcript
“If you make it cheaper to do something what happens do you do the same for less money or do you do more for the…”
“before Excel junior investment bankers worked really long hours and now thanks to Excel Goldman's associates all work at lunchtime on Fridays”
“to begin with you do the old thing but more with every new technology you do the old thing but more of it on the…”
“what does this change what wasn't possible before what gets unlocked as opposed to just doing the old thing but more of it”
From the episode
A rational conversation on where AI is actually going
Benedict Evans