The Lubricant Strategy: Why Top Students Fail Interviews (And How to Stop)
TL;DR: Your exam skills — memory and reasoning — are not worthless in the real world. They're the exact skills that get you hired, promoted, and paid. The problem isn't what you have. It's that you forgot to show it.
James here, CEO of Mercury Technology Solutions. Hong Kong — August 2026
A reader asked me a brutal question.
He'd read my piece on how the entire exam system really only tests two things: memory and computational reasoning. Everything else — creativity, charisma, street smarts — that's on you to develop.
His problem? He had nothing else. Just those two skills. And he was sitting in an interview tomorrow, acutely aware that he had no "real value" to offer.
How do you bluff your way through with nothing but exam skills?
The answer surprised him. It might surprise you too.
The西游记 Illusion
Here's what nobody tells you about companies above a certain size:
They're all Journey to the West.
You see four pilgrims walking to India. But remove Zhu Bajie and Sha Wujing — does the mission fail? Of course not. Sun Wukong carries the entire operation. Tang Sanzang is just there to get captured.
Bill Gates understood this in the 1990s. He said if he could take 100 people out of Microsoft, that would be Microsoft. The rest? Replaceable.
Elon Musk proved it decades later. He fired 85% of Twitter's staff. The site kept running.
Most employees are not critical parts. They're lubricant.
Your goal isn't to be irreplaceable. Your goal is to secure your position in the 80% that keeps the machine running smoothly — and get promoted before the next round of cuts.
The Two Skills That Actually Matter
Remember what the exam system drilled into you?
- Memory
- Computational reasoning
These aren't useless. They're exactly what corporate environments reward. You just need to know where to apply them.
Memory as Power
Here's a scene from 2009. I was an architect at a software company. Three departments, fifteen people, three hours into a meeting about a bug that was going nowhere.
They were preparing to extend into overtime. I had friends waiting at dinner place.
I stepped out. Fixed the code. Committed the patch. Fifteen minutes total.
Came back and said: "Stop arguing. The problem doesn't exist anymore."
Here's the thing — everyone in that room could have done what I did. The difference wasn't capability. It was that I actually did it while they were still debating whose responsibility it was.
But that's not the point. The point is what happens after you join.
The Meeting Room Is the Real Battlefield
In any mid-to-large company, most of your job isn't building things. It's meetings.
People talking over each other. Someone silent in the corner — you can't tell if they're thinking or asleep. The same arguments on loop.
Sound familiar?
This is where memory becomes your superpower.
A good memory lets you do something magical in meetings: cite specifics.
"On March 15th, the backend team pushed commit a7f3d2e that changed the authentication flow. That's when the latency spikes started."
You don't need to have written that code. You just need to remember it.
The director in the room doesn't read your pull requests. He doesn't know who fixed what. But he watches meetings. And when you start dropping dates, commit hashes, and module names from six months ago like you're reciting poetry, one thought forms in his mind:
This person knows everything.
That's not manipulation. That's information asymmetry working in your favor.
Reasoning as Translation
Memory gets you noticed. Reasoning gets you promoted.
Most people in meetings can describe problems. Very few can explain the logic in a way that decision-makers understand.
When the director is lost in technical weeds, someone needs to translate:
"What they're saying is: the database write is blocking the read queue. We have two options — shard the writes or implement async processing. Option two is faster to deploy but harder to maintain."
You don't need to be the top engineer. You need to be the person who makes the top engineers' work legible to power.
That's how you end up leading people smarter than you.
The Interview Is Just a Meeting With Higher Stakes
Back to the reader's problem. He's in an interview with "no value."
Wrong frame.
The interview is a meeting where you haven't been hired yet. The same rules apply.
Before you walk in:
- Memorize everything publicly available about the company
- Know their products, their competitors, their recent news
- Understand the department's role in the org chart
You're a top student. Memorizing information is what you were trained for.
During the interview:
- Speak in structured logic: premise → evidence → conclusion
- Cite specifics: "I noticed in your Q2 report that cloud revenue grew 34%..."
- Show reasoning: "That suggests you're shifting from on-premise, which means this team probably needs someone who can bridge legacy and cloud architectures"
You don't need to be the best candidate. You need to be the candidate who appears to have the clearest mental model of their business.
The Hard Truth (And the Warning)
Let's be honest: most people in most companies are not exceptional.
They're competent. They show up. They move papers (or Jira tickets) from one column to another. And many of them become managers, directors, even VPs.
How? They understood the game.
The game is visibility, not virtue. The people who get promoted are the ones decision-makers see thinking clearly and speaking precisely. Not necessarily the ones doing the best work in the dark.
If you have memory and reasoning skills — and most top students do — your only failure mode is forgetting to demonstrate them.
But here's my warning, and I mean it:
Don't get comfortable.
When I was building my career, we were in an expansion phase. Companies hired aggressively. They trained fresh graduates. There was room to grow into your role.
You're not in that world.
Twitter's 85% cut isn't an anomaly. It's a preview. As AI handles more execution work, the "lubricant" class shrinks. The meetings where you could demonstrate your memory become rarer. The translation layer where your reasoning shines gets automated.
南郭處士 could hide among 300 musicians. Try hiding when it's one human and 299 AI agents.
The Real Strategy
Use your exam skills to get in. Use them to get noticed. Use them to buy yourself time.
But while you're doing that — build the other skills.
The ones no exam tests:
- Judgment under uncertainty
- Building relationships across power structures
- Creating value that doesn't require a meeting to validate
- Developing taste for what matters versus what's merely urgent
Your memory and reasoning got you this far. They're enough to get you through the door.
But the door is getting narrower every year.
Don't be the person who optimized for a game that's about to change rules.
Mercury Technology Solutions: Accelerate Digitality.
