IDEAS / POST
Are you creating something of value, or just ‘hacking a test’?
What Paul Graham’s “The Lesson to Unlearn” teaches us about how to create value instead of just performing competence
Paul Graham’s essay “The Lesson to Unlearn” is a brilliant diagnosis of how school trains us to win by hacking bad tests, and how that thinking carries into how we build companies, products, and strategies. Jacob Beckerman, CEO of Macro, recently pointed me to it.
Graham’s thesis: we’re conditioned from an early age to optimize for the proxy (the grade) instead of the thing it’s supposed to measure (actual learning).
In business the proxies change, becoming awards, visibility, stakeholder praise, and funding announcements. The reflex doesn’t. We chase the score instead of building something of real value.
This isn’t just a personal development insight. It’s a strategic blind spot for many leaders and teams.
Here’s how the lesson to unlearn plays out in the wild
Startups obsess over getting funding, instead of solving painful customer problems.
Brand builders chase impressions and awards instead of revenue and retention (ummmm, I’ve done this on more occasions than I’d like to admit. The entire ad industry is built on it).
Internal teams pitch transformation projects designed to win board approval, not make the company better.
This mindset is deeply embedded. It’s not laziness. It’s learned behavior.
Remember when it dawned on you that you’d ace a medieval history course not by learning medieval history, but digging into the bits our prof belaboured? Cherry pick the vanity metrics… and ignore the real point of learning.
If this wrongheaded and ultimately counterproductive tic isn’t surfaced and unlearned, it creates a gravitational pull toward busywork disguised as progress.
The lesson to unlearn, from Y Combinator
Young founders would constantly ask: How do we get VCs to invest in us? What’s the trick?
And Graham would tell them something that should have been obvious: Be a good investment. Build something people want. Make the product great.
But instead of focusing on improving their product or solving real user problems, they’d gravitate to superficial hacks:
- How to time their launch for maximum exposure to investors (always on a Tuesday),
- Which influential people they should try to get introductions to,
- How to craft a pitch deck that hit all the right buzzwords.
These are all relics of bad-test thinking. A belief that there’s a trick to winning, a shortcut, a cheat code. That if they could just hit the right formula, they’d get the grade. But building a startup, like building any great business, isn’t about tricking the test. It’s about doing good work.
That simplicity is radical. And it’s exactly the pivot I help clients make.
At the heart of every high-performing company, product, or brand is something elegantly simple: a core idea, a point of view, a thing that it does uncommonly well and with undeniable usefulness. But more often than not, teams bury that essence under layers of well-intentioned optics, process, and second-guessing – hacks to quick success.
My work is about cutting through that complication. Helping leaders return to the real engine of value in their business. Not the metrics they’ve been told to chase, or the formats they’ve been trained to deliver, but the substance that makes their company singular and attractive.
As an advisor, I see my role as helping clients
- Recognize when they’re optimizing for a hackable test (investor sentiment, internal KPIs, perception management).
- Reorient toward unhackable measures of value: user behavior, economic outcomes, strategic asymmetry.
- Say no to performative work, even if it looks right.
The shift is powerful. When clients stop trying to pass a test and start trying to build something real, energy floods back into the work. Decisions get clearer. Teams move faster. Outcomes improve.
So if you’re in a position of leadership, ask yourself
Where in your organization are you optimizing for a test rather than the actual goal – building a better company?
If this hits close to home, I’d love to hear how you’re unlearning bad-test thinking, or helping others do it.
Frequently Asked Questions
What is Paul Graham’s “The Lesson to Unlearn” about?
Published in December 2019, the essay argues that school teaches us to win by hacking bad tests, meaning we optimize for grades rather than learning. That habit follows people into adult work, including building startups, where it pulls attention toward tricks instead of doing good work.
What does “hacking a bad test” look like in a company?
It means optimizing for a proxy that’s supposed to signal value instead of creating the value itself. Common examples are investor sentiment, awards, impressions, board-friendly transformation projects, and internal KPIs that can rise while customer behaviour stays flat.
What did Paul Graham tell Y Combinator founders who asked how to raise money?
He told them the best way to get VCs to invest is to actually be a good investment. What makes a startup a good investment is growth, ideally in revenue and otherwise in usage, and the way to get users is to make the product really great. The founders kept going back to launch-day exposure, influential introductions, and launching on a Tuesday.
How can founders tell if they’re optimizing for a proxy instead of real value?
Graham’s rule of thumb is that tests imposed by an authority are usually proxies for something else, which makes them hackable. A practical check: if this metric improved and customer behaviour didn’t change, would the business actually be better off? If the answer is no, you’re working on the test.
What are unhackable measures of value for a growing company?
Measures that customers control rather than evaluators: retention, repeat purchase, referrals, revenue and margin, and strategic advantages competitors can’t easily copy. These are hard to game because they only move when the product and positioning genuinely work.