Real talk from a technical founder building AI-powered businesses
Published September 30, 2025 • Based on Founder Reality Episode 23
Also available on: Apple Podcasts • Spotify • YouTube
Three stories caught my attention this week. When you connect the dots, they reveal something fundamental changing about how you actually build businesses in 2025.
The old playbook is dead. Raise big VC rounds, scale fast, worry about revenue later - that approach worked for the past 15 years, but it's not working anymore. And the data proves it.
Let me show you what's actually working now.
Story 1: Founders Are Walking Away From Traditional VC (And It's Not Because They Can't Raise)
Mercury just surveyed 1,500 early-stage startups about how they're funding their companies in 2025. The results challenge everything we thought we knew about startup funding.
Here are the numbers that matter:
- 66% of founders changed their capitalization strategy in the past year
- Only 22% of companies raised under $1M in their last round
- 73% raised under $5M total
- Companies using 4+ different funding sources are 40% more likely to raise $5M rounds
- 61% of companies rely on contractor talent, not full-time employees
Read that last stat again. The majority of startups aren't hiring full-time employees anymore. They're using contractors.
This isn't about founders getting desperate. It's about founders getting strategic.
The Consulting-First Approach Nobody's Talking About
I've lived this personally. In 2022, I tried to raise VC funding. Almost closed a round, but it fell through. Instead of chasing more meetings, I bought the startup with my own money and decided to be revenue-first.
After about a year of self-funding, we hit break-even and became profitable. It's not just about not giving away control - it's about having more options for how to build your business.
Here's the approach I've been recommending to founders (and I'm releasing a free ebook about this soon):
Don't start by building a SaaS product. Start with an idea, bring it to ideal customers, and sell a service instead of a product.
Why? Because I've done this for three years and it works.
When you sell a service, customers understand it immediately. They're not wondering about onboarding time, employee buy-in, or learning curves. They know they can rely on you to do the work for them.
The new funding mix founders are actually using:
- Consulting revenue - Immediate cash from delivering services
- Grants - Non-dilutive funding that doesn't take equity
- Strategic partnerships - Companies pay you to deliver services
- Small angel checks - Easier to close than VC rounds
Your first $100K should come from customers, not investors. Once you have revenue, everything else becomes significantly easier.
Story 2: Perplexity Got Copied By Every Big Tech Company (And They're Still Thriving)
In December 2022, Perplexity launched their answer engine - AI search that gives you direct answers with real-time web crawling and citations, instead of linking to Google.
Then everyone copied them. ChatGPT added search. Claude added search. Google went aggressive with AI overview and AI mode. Gemini added search features.
The Perplexity CEO told founders at a startup school: "If your company can make revenue on a scale of hundreds of millions of dollars, you should always assume big companies will copy it. They will copy anything that's good and you got to live with that fear."
Here's the plot twist: Perplexity didn't die. They're still growing.
I was actually pessimistic when they launched their browser. I thought "who's going to use an AI browser?" But now I'm interested. I know people who subscribe to Perplexity - which is huge for a simple AI solution.
They survived because they didn't try to out-engineer Google. They stayed true to their original mission:
- Make it load really fast (fastest I've seen)
- Give answers immediately with the fastest throughput
- Build a brand around being anti-Google for people tired of clicking blue links and reading SEO-optimized garbage
They competed on experience, not technology. Instead of spending billions training their own LLM, they built in-house models based on open source and optimized for speed.
The Green Sky Lesson I Learned The Hard Way
When I started SimpleDirect, I obsessed over our competitor Green Sky. They went public about a year after we started and hit around $10 billion valuation. I thought "what the hell, our competitor is doing so much better, let's copy what they're doing."
That was the wrong approach.
Green Sky didn't succeed in their game. They got merged and sold off multiple times after being taken private. OpenSax acquired them and lost a ton of money.
The lesson: Don't copy what your competitors are doing. If you're a startup, there's a reason people use you instead of the competitor. Find that reason and double down on it.
For Perplexity, they're not just "AI search." They're "AI search for people who hate how Google works." That positioning cannot be copied.
Story 3: The ARR Theater Problem That's Hurting Honest Founders
Fortune magazine just published an investigation about founders abusing the term ARR (Annual Recurring Revenue).
Here are some actual examples:
- Clueless (controversial startup) claimed to double their ARR from $3.5M to $7M in one week
- Unnamed startup claimed $325K in ARR, but when VCs dug deeper, it was just a two-week pilot contract with a handshake deal
The patterns VCs are seeing:
- Counting pilot programs as recurring revenue
- Claiming "booked ARR" based on contracts with exit clauses
- Treating customer possibilities as guaranteed income
VCs are calling this "vibe revenue" - and they're right.
Why This Makes Me Angry
This hurts legitimate founders who are actually building sustainable businesses. When honest founders report real ARR now, VCs are skeptical and demand more documentation.
I'm very careful with revenue numbers. I use MRR (Monthly Recurring Revenue) for SimpleDirect because it's an accurate representation of where we are. For ARR, I only count actual recurring revenue from the previous full year.
Some startups launch, get a few customers, multiply their revenue by 12, and call it ARR. That's borderline acceptable. But having pilot programs with sketchy contracts where customers haven't even paid yet and calling that ARR? That's just wrong.
The Pressure Behind The Numbers
I understand why it's happening. Everyone cares about revenue now in 2025. Founders want to show growth to raise money. Everyone's trying to be the next Cursor (zero to $100M ARR in one year).
But here's the thing: Cursor worked because they were first and they were better. They beat GitHub Copilot, ChatGPT, and Claude at what they do. I use Cursor sometimes and I love it.
Trying to take shortcuts with inflated numbers doesn't make you the next Cursor. It just makes you dishonest.
How VCs Should Actually Evaluate Revenue
If you're a VC or investor, look at three categories:
- Locked revenue - Signed contracts with money in the bank
- Probable revenue - Strong pipeline with clear next steps
- Possible revenue - Everything else
Count locked and probable for real MRR/ARR evaluation. Ignore the inflated numbers.
The New 2025 Playbook
The old playbook:
- Raise VC money first
- Build fast and scale aggressively
- Throw money at marketing, sales, ads
- Worry about competition later
- Show hockey stick growth and raise more rounds
This worked 15 years ago for Uber, Airbnb, DoorDash. But we're in a different time now.
The new playbook:
- Build revenue streams that cannot be copied - Through relationships, content authority, positioning, and speed
- Use ecosystem approach - Build adjacent products that make your ecosystem stronger
- Build community - Loyal following that relies on you and your business
- Stack different funding sources - Grants, early revenue, consulting
- Report honest numbers - Don't inflate metrics, build credibility instead
My Prediction for 2027
The companies thriving in 2027 will be the ones that:
- Started with consulting revenue
- Built strong customer relationships and community
- Used those as foundation for business decisions about funding and competition
VC funding will still be important, but it won't be the only thing startups count on.
Your Action Items This Week
- Audit your revenue reporting - Are you reporting real, locked revenue or "vibe revenue"?
- Think about your defensibility - What happens if big tech copies your product tomorrow?
- Consider the consulting approach - Can you sell your idea as a service before building the product?
- Build community - Start creating content and authority in your space
The game has changed. The founders who adapt to this new playbook will be the ones who survive and thrive.
The Bottom Line
Stop chasing the 2019 playbook. Stop inflating your metrics. Stop trying to be the next unicorn by copying what worked 10 years ago.
Start with real revenue. Build real relationships. Create real value that can't be copied.
That's how you build in 2025.
Want my free ebook on the consulting-first approach to building startups? Coming soon at founderreality.com
Got news stories or thoughts on these topics? Email me at george@founderreality.com or DM me @TheGeorgePu on Twitter - my DMs are always open.
New episodes of Founder Reality drop Monday/Wednesday/Friday at 9am EST. Real founder insights from the trenches, not startup theater.
