Key Principle
A platform's value is mostly created by people the maker never employs. Two decisions made the Apple II a platform rather than a finished gadget: Wozniak's eight expansion slots (over Jobs's objection), which let third parties extend the machine, and VisiCalc, the "killer app" that inverted sales logic so that software demand pulled hardware sales. Apple invented neither — its discipline was converting an accidental software advantage into a moat (distribution exclusivity) and subsidizing rather than strangling the outside developers who built most of the value.
Why This Matters
The hardest truth about platforms is that the maker is not in control of why people buy. "People bought the computer to get the software, not the other way around" (The Bozo Explosion). The fate of a platform can be decided by an application its maker "never built and barely understood" — and the people closest to the product systematically misjudge what the market will buy. The strategic question is therefore not "what should we build?" but "who else can we let build value on us, and how do we keep our advantage while they do?"
The slots fight shows the bet is strategic, not technical. Capping extensibility (Jobs wanted two slots) would have foreclosed the peripheral and software ecosystem that actually drove the Apple II's success. Extensibility is a strategic bet on third parties you cannot foresee.
Good Examples
- Eight slots over two. Wozniak borrowed eight expansion slots from minicomputers; Jobs argued "you would never need more than two." Wozniak refused to let it go with two. The slots "brought benefits to everybody" — the manufacturer got a subindustry to boast about, peripheral makers got new products, customers got one machine for many jobs (A Lot of Poop).
- VisiCalc as the tail that wags the dog. Analyst Ben Rosen framed it: VisiCalc "could someday become the software tail that wags (and sells) the personal computer dog" (The Bozo Explosion). It ran only on Apple for ~12 months after its Oct 1979 launch ($100); Scott estimated 25,000 of the 130,000 machines sold before Sept 1980 (~19%) moved on its strength.
- Subsidize, don't strangle. Open slots plus programmer discounts created an external R&D army (80-column cards, modems, Microsoft's CP/M Softcard, Easywriter). Apple bankrolled shoestring developers — Budge traded his game Penny Arcade for a $1,000 printer; Gibbons collected his Apple at Jobs's home (The Bozo Explosion).
Counterpoints
- The advantage was a coin-flip, not a plan. Bricklin built VisiCalc on an Apple only because the Commodore and Radio Shack loaners were in use; he "might as well continue" (The Bozo Explosion). The discipline was in productizing the accident via exclusivity, not in inventing it.
- Proximity is not insight. Markkula and Atari read the VisiCalc demo as "a checkbook program"; the outsider analyst Rosen grasped it first. Insiders misjudge their own platform's killer app.
- Closing the platform backfired (Apple III). Having won on third-party software, Apple tightened technical secrets and tried to build software in-house, making it "almost impossible for independent software companies to develop programs" — starving its own machine of the exact thing that made its predecessor valuable (Chunk 024 — Founder-as-Brand & the Apple III Failure).
Key Quotes
"All Steve saw was a computer that could do a couple of things... you would never need more than two slots. I refused to let it go with two slots." — Wozniak, Chapter: A Lot of Poop "Visicalc could someday become the software tail that wags (and sells) the personal computer dog." — Ben Rosen, Chapter: The Bozo Explosion "There is no one who sells as well as a committed, involved user." — Chapter: The Bozo Explosion "Software is the glass through which the majority of our users see the Apple. If it doesn't work right, the Apple isn't working right." — Jef Raskin, Chapter: The Bozo Explosion
Rules of Thumb
- Design for extension, not just for the use cases you can see today — let the builder's platform instinct override the merchant's "you'll never need it."
- Treat your killer app as a moat to defend (exclusivity, distribution), even when you didn't write it.
- Subsidize outside developers (discounts, hardware, hand-holding); they are unpaid R&D and your most credible sellers.
- Distrust insiders' read on what the market wants from your platform; the decisive insight often comes from an outsider.
- Never close a platform that won on openness — proprietary control becomes self-sabotage.
Related References
- Productize, Don't Invent - converting others' work into product, not inventing it
- The Builder + Merchant Partnership - the builder (Woz) holding the platform vision over the merchant (Jobs)
- shipping discipline - the Apple III's closed-platform reversal as failure mode
- Success Breeds Arrogance & Underestimating Incumbents - antagonizing the developers who made the platform valuable