Start-ups and AI as an experimentation machine

AI is everywhere. And it is a perfect start-up tool. Particularly in helping  answer the same fundamental questions start-ups always have: 

• Who is my customer? 
• What problem can I help them solve? 
• How do I make money and build a sustainable, profitable business?

“The Experimentation Machine: Finding Product-Market Fit in the Age of AI”

According to “The Experimentation Machine: Finding Product-Market Fit in the Age of AI”, an AI is the perfect Experimentation Machine. With scientific rigor. Accessing the best minds in the business. A startup founder’s job is that of Chief Experimentation Officer.  Startups are businesses without working business models. 

Every business is comprised of the same parts: a customer value proposition, a go-to-market plan, a technology/operations infrastructure, and a monetization strategy (ideally one that eventually makes the business a profit). They are discovering and creating entire elements of the business model on the fly, often in a messy fashion.

A successful startup needs the right mix of founder-market-fit, investors, team members, and channel partners. None of these stakeholders are guaranteed in the early days. So how on earth does any startup get off the ground? By experimentation. And AI can help there in a big way.

Test the market

The goal of a start-up is to test and probe for answers to the most fundamental questions of any business: Who is my ideal customer? What do they want and need? How do I deliver it to them? How do I find more ideal customers? How do I make money and build a sustainable business?

After conducting deep customer discovery—the process of thoroughly researching potential customers’ needs and problems—the founding team creates a hypothesis or an educated guess.

Founders need to make hypotheses for every part of their business: customer value proposition, go-to-market, technology infrastructure, and monetization plan. A hypothesis must be falsifiable. Can you definitively prove it to be correct or incorrect? Then you identify the most important hypothesis and design an experiment to test

Finally, founding teams must analyze and implement the results of the experiment.

The journey

The journey through Startupland goes through three distinct phases:

• The Jungle: The land is wild and dangerous. There are no paths and no set direction. 
• The Dirt Road: After trial and error, you eventually clear a way through the jungle.
• The Highway: Through sheer effort and a little bit of luck, you pave your dirt road into a highway,

Test selection as a strategy

The strategy of test selection boils down to three essential questions:

• Which business model component is most controversial, and what is the essential hypothesis for that component?
• What is the key milestone I need to achieve that will lead to my next valuation inflection point, helping unlock more capital from investors?
• Where does the greatest risk exist in my business model, and what does the flow of dependencies look like?

Finding the fit

There are 4 levels of product-market fit:

• Nascent: You have a handful of somewhat engaged and happy initial customers, but things still feel early and messy.
• Developing: You have more engaged, paying customers and less churn—you need to work on driving demand.
• Strong: Momentum is picking up, and you’re finally feeling the “pull” of demand. It’s time to focus on increasing efficiency.
• Extreme: You’re repeatedly and efficiently solving an urgent problem for a large number of customers who need your product.

HUNCH

The author has a framework to test that called HUNCH:

• Hair-on-Fire Customer Value Proposition
• Usage High 
• Net Promoter Score 
• Churn Low 
• High LTV:CAC Ratio

For that framework alone, the book is worthwhile. Jeff Bussgang knows what he is talking about. He has a few more:

The go-to-market hypotheses

There are four questions to answer:

1. What is my initial market? 
2. What is my growth and demand generation strategy? 
3. What is my sales model? 
4. Who are the best (if any) channel partners?

Understanding startup valuations

There are four factors to consider:

1. A strong competitive moat.
2. Recurring revenue.
3. Strong network effects.
4. Attractive unit economics or unit profitability.

RAWI

• Ready? Are you seeing strong demand for the product, and are you satisfying customers? •
• Able? Do you have access to the necessary human, capital, and technology resources required to scale successfully—that is, is the team in place, and can capital be raised?
• Willing? Are you committed to growing the business, despite the risks of scaling, and will growth advance your original vision?
• Impelled? Does the startup have aggressive rivals, and is the market characterized by strong competitive intensity? Is speed and scale imperative to success?

Technical debt

There are four versions:

1.  Deliberate and reckless debt happens when a team knowingly cuts corners to get quick results, which often causes problems later for startups.
2. Deliberate and prudent debt occurs when a team intentionally pushes off some work for strategic reasons and plans to fix it later.
3. Inadvertent and reckless debt comes from not knowing good design practices, leading to messy code that slows progress.
4. Inadvertent and prudent debt is the natural result of learning during a project.

Scaling go-to-market: sales, marketing, and expanding TAM

The stages:

Stage 1: Founder selling: In the earliest days of your business, sales should be led by a founder. There is no one on the planet with as much passion or insight into the specific problem you’re trying to solve.

Stage 2: Renaissance reps: A startup will eventually hire a salesperson. But whom?

Stage 3: First head of sales: The mission of the first head of sales is to make improvements to whatever sales system the founder-seller set up in the first two stages of the sales learning curve.

Stage 4: Coin-operated selling: This stage is called “coin-operated” selling: sales reps are highly motivated by financial rewards to close deals and hit their numbers, and they do so repeatably and predictably.

Scaling

• Horizontal scaling: This involves expanding into new markets or segments with the same or a similar product, platform, or service.
• Vertical scaling: This approach focuses on deepening penetration within a specific industry, market, or customer segment by adding new products and services.
• Geographic scaling: This growth vector involves expanding into new geographic markets.
• All of the above

Blitzscaling

The author refers to blitzscaling. A strategy for rapidly scaling a company to achieve market dominance. I am not a fan. Wrote a blog about it here.

Other tips

• Pick an initial market that is consistent with your passion and mission.
• Select a market that is narrow enough to ensure focus, but large enough to attract capital and sustain multiple iterations.
• Identify a consistent set of features across customers.
• Your initial market should demonstrate a high willingness to pay.
• Ensure you have direct access to your customers (i.e., no gatekeepers).
• A LTV:CAC ratio should ideally be 3:1 or higher.
• Do not forget to factor in the gross margins
• Do ot underestmate churn rates.
• Factor in the payback period.
• Do not underestimate the costs of technical debt.
• Hire for jobs-to-be-done, not titles
• Always focus on 10X

Moving faster with AI

AI is the greatest leverage-booster for startups since the Internet. We are entering a new era of startups, where founders are accomplishing much more with much less, rapidly accelerating the search for product-market fit. At a 10X rate.

AI is perfect for

• Hypothesising: target customer, problem definition, solution, etc.
• Comprehensive market analysis: summarizing, synthesizing, and recognizing patterns.
• Competitive intelligence: crawling the web and ingesting publicly available data, they are masters of competitive intelligence.
• Problem discovery: brainstorming partner for problem discovery.
• Create hyper-specific customer personas: refining and bringing to life customer personas, tailoring product features and marketing strategies to match identified customer segments.
• Saving them time on manual prospecting, customer research, and outreach.
• Personalization at scale.

Startups as a science 

AI does allow you to embrace what the book calls your Inner Edison and develop an experimentation mindset where you approach your venture like a scientist. Every new idea or strategy starts as a testable hypothesis.

Going faster

Artificial intelligence also unlocks startups from the shackles of capital and human headcount to grow faster and leaner than ever. But it’s also a powerful and sometimes unwieldy tool that founders must learn to harness.With AI copilots and agents, startups can now build well beyond an MVP-level prototype to scale their product without amassing an enormous engineering budget.

Strategy is still human

My problem with AI is that a plan works on paper until you get boxed in the face. AI does not change that. Everything is easy in a sandbox. However, AI is dramatically expanding our capacity to run experiments and build products, but startups still face time and resource limitations, and thus need to uncover the right answers quickly. However strategy is old-fashioned human. As artificial intelligence takes over more of the what and even the how, humans must become experts in the why. Read this. 

Do not wait to use AI

The main message of the book?  Do not wait. There is no escaping AI. Whatever AI you’re using right now is the worst AI you’re ever going to use. AI may not replace founders. But founders who learn how to use AI effectively will outperform founders who don’t. Founders who fear they’ve already missed the AI train should recruit a few “reverse mentors” to teach them—perhaps someone less experienced than you, but already facile with AI.

PS

Also read “Artificial Organizations: Build Better Judgment, Speed, and Results with Human and Machine Intelligence”. They fit perfectly.