One-page summary · 2 min read
The Innovation Stack
Building an Unbeatable Business One Crazy Idea at a Time
by Jim McKelvey
The big idea
Jim McKelvey argues that durable, defensible companies aren't built on a single innovation but on an innovation stack, an interlocking set of unplanned solutions a company is forced to invent, in sequence, to survive problems with no existing playbook.
When an entrepreneur tackles a genuinely new problem rather than copying an existing business, solving it typically creates a new problem needing its own solution, and so on. Each layer depends on the ones below it, so a competitor can't simply copy the top layer, they'd have to independently rebuild the entire stack. McKelvey illustrates this with Square's founding: a lost sale because he couldn't accept Amex led to building a card reader, which ran into 17 different credit-card-network regulations, each requiring its own novel workaround.
McKelvey frames the stack as something that emerges organically under survival pressure, not something a company can deliberately plan from the outset.
Key ideas
Perfect problems
The best entrepreneurial opportunities are solvable problems with no existing solution, so there's no playbook or competitor to copy.
Innovation as involuntary
True innovation stacks rarely arise from deliberate choice, they're forced into existence because a founder is blocked at every turn with no incumbent solution to fall back on.
A.P. Giannini and Bank of Italy
McKelvey uses banking pioneer Giannini, who served immigrant and working-class customers ignored by big banks, as a historical example of an innovation stack built long before the term existed.
Timing and enabling technology
Some innovations only become possible once underlying technology matures enough to support them.
Constraints as an advantage
Operating outside an established market means lacking conventional resources, but this absence of how it's normally done can free founders to build genuinely different solutions.
Stacks as competitive moats
Because a full stack is hard to replicate piece by piece, it can protect a company from being out-competed by larger, better-resourced incumbents.
Try it this week
- Write down a problem in your work with no existing solution you can copy, and list every sub-problem solving it would create.
- Identify a workaround you use only because no proper tool exists yet, and treat it as the first layer of a potential stack.
- Look at a competitor's most defensible advantage and ask whether it's a single feature or an interlocking stack of decisions.
- Map the Square and Amex origin story against a frustration you've personally had as a customer, is it a generalizable problem?
Is it for you?
Best for
- Founders and product leaders building something without an existing playbook
- Entrepreneurship-minded readers wanting an alternative to copy-and-iterate startup advice
Less useful if
- Readers seeking a repeatable, teachable methodology, since McKelvey's own argument is that stacks can't be engineered on demand
- Managers inside large, established companies looking for internal-innovation tactics rather than founder-level survival stories