Core-Competency Outsourcing ("Beer Test")
If it doesn't make your beer taste better, outsource it.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 70%
Referencing Jeff Bezos's framing for AWS, the hosts articulate a simple filter for resource allocation: if an activity doesn't directly make your core product better for the customer, it should be outsourced rather than built in-house. The flip side is equally important -- when you notice many companies all outsourcing the same non-core function, that pattern itself is a signal of a platform business waiting to be built (which is effectively what TSMC became for fabless chip companies). The framework is a quick, repeatable filter for founders and operators deciding where to spend scarce engineering and capital resources.
Origin
Popularized by Jeff Bezos's German-beer-factory analogy for why AWS outsourced infrastructure like electricity generation for early breweries.
Core principles
- 01Differentiation, not importance, determines what should stay in-house
- 02Every function that doesn't touch the core differentiator is a candidate for outsourcing
- 03Widespread industry outsourcing of a function signals a platform opportunity for someone else
- 04Freed resources should be reinvested directly into the core differentiator
How to run it
- 1
Define What Makes the Beer Taste Better
Identify the specific one or two activities that are your true differentiated value-add to the customer.
- 2
Audit All Other Functions
List every other operational function -- finance, manufacturing, infrastructure, HR -- that doesn't directly touch that differentiator.
- 3
Outsource or Buy Rather Than Build
For every function outside the core, default to buying or outsourcing from a specialist rather than building in-house.
- 4
Watch for Mass-Outsourcing Patterns
When many companies in an industry are all outsourcing the same non-core function, treat that as a signal of a platform opportunity to build for them.
Pro tip This is exactly how TSMC turned fabless companies' shared weakness into its own platform business.
- 5
Reinvest Freed Resources into the Core
Redirect the capital and attention saved from outsourcing back into strengthening the actual differentiator.
In the wild
Fabless companies like Nvidia, Qualcomm, and Apple all outsource chip manufacturing -- not their differentiator -- to TSMC and focus their own resources on chip design.
→ Nvidia reached hundreds of billions in market cap after raising only $20 million, without ever building a fab of its own.
Common mistakes
Confusing 'Important' with 'Core'
Manufacturing is critically important to a chip company's success, but that doesn't mean it needs to be done in-house; conflating importance with differentiation leads to unnecessary capital lock-up.
Is it for you?
Best for
Startups and lean teams trying to focus limited resources on their true differentiator
Not ideal for
Companies where the 'non-core' function is actually becoming a strategic differentiator (e.g., proprietary infrastructure at hyperscale)
From the transcript
“As a startup, anything that doesn't make your beer taste better, the analogy back to German beer factories and outsourcing electricity generation. Outsource things that…”
“The counterpoint to that is anytime you see something that lots of people, lots of companies are doing that is not making their beer taste…”
From the episode
TSMC