How 80% of my revenue disappeared twice in 60 days because I built on someone else's infrastructure. The brutal lessons about middleman businesses, the four principles for building anti-fragile companies, and the three-question audit that reveals if you're one policy change away from disaster.

The double shock that almost crippled both businesses:

• Two massive shocks hit SimpleDirect and ANC within 60 days
• 80% of inbound leads vanished for both businesses simultaneously
• Not because products got worse, not because of competitors
• Because third parties I had zero control over changed the game
• Almost crippled my income overnight

Currently living through this (70-80% to the other side):

• Both businesses pivoting right now - websites changing, value props changing
• Still figuring things out, rebuilding, getting stronger
• Sharing while living through it, not after everything's solved
• This is raw, real, and happening to more founders than want to admit it

ANC story: When government changes the rules:

• Consulting business helping international students establish businesses in Canada
• Focused on Start-Up Visa pathway - 10-15 premium clients yearly
• Strong margins, real transformations, happy clients
• But the pathway (controlled by government) was the hook bringing people to us
• Government approval timeline: used to be 12 months, now stretched to 85-87 years
• Not a typo - EIGHTY-SEVEN YEARS for approval
• No warning, no announcement - just pulled levers behind scenes
• 80% of our leads evaporated as rules kept changing
• Built valuable service on trap door someone else was holding

SimpleDirect story: When lending partners control your fate:

• SimpleDirect Financing was flagship - connected contractors to lending marketplace
• One application, 10+ banks/fintech lenders, best rate matching
• Relied solely on lending partners for approval rates
• Customers came for financing results, not because they loved our product
• When lenders' APIs went down or made bad decisions, customers blamed US
• Customer support tickets became overwhelming
• Shutting down SimpleDirect Financing December 31st, 2025
• Launching ChangeLock - product we control end-to-end with no third-party dependencies

The pattern I missed twice (same mistake, two businesses):

• I was facilitator, not builder
• Helped people navigate someone else's system
• Expertise was valuable, transformations were real - but didn't own the outcome
• Was coordinator/customer experience wrapper, not actual infrastructure
• Felt safe at first - had revenue, getting paid, business working
• Lower upfront investment, faster to revenue - made total sense
• Until it didn't

Why you can't diversify fast enough when primary channel dies:

• Building new infrastructure takes 12-18 months minimum
• When something dies, you're digging yourself out of hole - doesn't work that way
• Runway burning, team stressed, customers confused, you're scrambling
• Even content marketing: can't launch Twitter tomorrow and get thousands of likes
• Need consistency and time for everything

Historical examples - infrastructure owner always wins:

• BlockFi: Billions in crypto lending, great UX, real value - vanished overnight when counterparty collapsed
• Travel agents: Knew everything about booking until airlines launched direct booking + Expedia happened
• Music labels: Controlled distribution until Spotify, YouTube Music, social media emerged
• Pattern: Middlemen create value initially, then infrastructure owners cut them out

The four non-negotiable principles for what I'm building now:

Principle 1: Own the transformation, not the transaction

• Old: "Come to us, we'll help you access this pathway/lending"
• New ANC: Transform founders $0 to $500K ARR with fundamentals so strong they qualify for 5+ options worldwide
• When one door closes, route to four others
• Transformation itself is the moat, not the pathway

Principle 2: Own the full stack

• SimpleDirect new: Build entire founder operating system (ChangeLock, Roadmap)
• Control pricing, UX, features, roadmap - no external dependencies
• Like Basecamp: build everything end-to-end, even own calendar and email client
• ANC new: In-person transformation experiences, not routing to someone else's program

Principle 3: Diversify ruthlessly

• ANC old: 80% leads from one source
• ANC new: Five sources, none over 30% - if one closes, four backups remain
• Product diversification: SimpleDirect (SaaS) + ANC (services) + equity in supported businesses
• Three revenue streams, three customer types
• Not about launching bunch of products - different business models that de-risk startup

Principle 4: Equity over transactions

• Take less cash, own piece of customers' companies
• Had chances to take equity in past, thought "we're consultants, not equity investors" - wrong
• Going forward: align incentives, make less transactional
• Own piece of customer experience and customer equity

The three-question audit every founder must run:

Question 1: What if they change the rules tomorrow?

• List every external dependency (platforms, APIs, partners, governments, suppliers)
• For each: "If they change terms tomorrow, would I survive?"
• If answer is no, you're a middleman
• Examples: Facebook ads CPM triples, supplier cuts margins 50%, approval process breaks

Question 2: Do I own the outcome?

• Do you deliver transformation/service end-to-end or coordinate someone else's delivery?
• Social media creators: if platform changes rules, traffic goes to zero
• George's mentor knew someone who jumped off building when Google algorithm change dropped web traffic to zero
• Don't build entire business on one thing you don't control - could be lethal

Question 3: Can I get cut out by infrastructure owner?

• Could customers go directly to your supplier/partner?
• SimpleDirect: customers could go to 10 lenders directly, we made money from information asymmetry
• In AI-first world, information arbitrage doesn't last
• If business hadn't failed for other reasons, this would have killed it eventually

The 30% Rule (critical for survival):

• No single dependency should represent more than 30% of revenue
• Not single customer over 30%
• Not single partner over 30%
• Not single platform over 30%
• Not single demographic/geographic market over 30%
• Violate this = danger zone

Example from The Anti-Unicorn book:

• Even at $10K MRR, if banking on 1-2 customers, you're in danger
• Need at least 5-10 different customers at $10K MRR before safe to quit job
• If rely on 1-2 customers, they can easily leave

What to do if you fail the audit:

Option 1: Expand your wedge

• Offer more pathways, product lines, features, partners
• Own more layers of your service
• Add service...