Overview of Founder-Led Sales to $1 Million ARR With Just 10 Customers
This episode features Felix Hoffman, co-founder and CEO of 7Learnings, a retail decision-optimization platform that helps brands and retailers make better decisions on pricing, marketing spend, and ordering. Felix explains how the company got started with a consulting project to access the data needed to build the product, how the first live pricing test was a disaster, and how they eventually reached $1M ARR with just 10 customers, all sold through founder-led sales.
The conversation also covers the company’s growth to $5M+ ARR, the role of A/B testing in proving value, the importance of customer patience in early enterprise sales, and Felix’s strong view that LLMs are not the right tool for deterministic enterprise decision automation.
What 7Learnings Does
7Learnings helps online retailers and brands make smarter, data-driven decisions by acting like a “Google Maps for retail businesses.”
Core use cases
- Pricing optimization: decide how to price products to maximize profit
- Marketing optimization: decide where and how much to spend on ads
- Ordering optimization: determine how much inventory to buy
Who it serves
- Mid-market to large retailers and brands
- Primarily online businesses
- Especially companies with enough volume for advanced optimization to matter
How the Company Started
Felix’s background in pricing consulting and later at Zalando shaped the idea.
Early validation path
- He saw that many companies were still making pricing decisions in a manual or outdated way
- At Zalando, he became interested in predictive decision-making
- He believed the approach could be generalized and improved for many retailers
The unusual first step
Instead of launching directly as SaaS, the company began with a consulting project:
- They needed a large, real dataset to train the system
- A consulting engagement gave them access to that data
- The client paid for consulting while allowing 7Learnings to use the data to build the product
This helped them develop:
- A forecasting engine
- An optimization engine
- An early front end for presenting recommendations
Founder-Led Sales and the First Customers
Felix said the first 10 customers came almost entirely through founder-led sales, and that this was essential in the beginning.
Why founder-led sales mattered
- External sales are hard to delegate early
- Enterprise buyers want trust, context, and direct access to the founder
- The first customers often need extensive handholding and explanation
What worked to acquire customers
- Personal network
- Industry events and conferences
- Customer referrals
- Some targeted outreach, though he noted this becomes less effective over time
Best-fit GTM approach
- More account-based than broad outbound
- Highly customized outreach to specific target accounts
- Speaking opportunities and masterclasses were especially effective
The First Pilot: A Disaster, Then a Win
One of the most memorable parts of the interview was Felix describing the first live deployment.
What went wrong
- The first price upload was a disaster
- The system priced high-margin products incorrectly and was far too expensive
- Because e-commerce feedback is fast, they realized almost immediately that it was not working
How they recovered
- They reworked the models
- They kept close communication with the customer
- They relied on a patient early adopter who understood the startup nature of the product
The eventual result
- The A/B test later delivered around a 13% profit uplift
- That kind of impact made the product valuable enough to support a high pricing model
How 7Learnings Prices Its Product
Felix explained that their pricing is built around value-based pricing, not performance-based commissions.
Pricing model
- A monthly fee
- The fee scales with the revenue being optimized
- Avoids complex success-fee structures
Why not success-based pricing?
- Creates extra pressure around A/B test results
- Increases complexity for customers
- Customers generally prefer simpler pricing structures
Key principle
If the product creates measurable profit uplift, the company can charge a fee that still leaves the customer with a strong ROI.
What Customers Objected to Early On
One of the main objections was that retailers were already price matching competitors and assumed that was the only reasonable strategy.
Common mindset Felix encountered
- “Everyone is doing it this way”
- Competitor crawling and price matching is seen as the default
- Many buyers don’t realize there are better ways to optimize pricing
Felix’s response
- Blind price matching ignores:
- margin impact
- stock levels
- seasonality
- demand shifts
- marketing effects
- The goal is not necessarily to stop following competitors, but to understand the consequences of doing so
Why Forecasting Matters So Much
Felix emphasized that the product depends on strong forecasting accuracy.
Why forecasting is essential
- You need to predict what happens if you:
- raise or lower prices
- increase ad spend
- discount heavily
- reorder inventory differently
- Better predictions lead to better decisions
Real-world examples
- Selling out during COVID or seasonal spikes
- Discounting products too early when inventory is constrained
- Over-ordering or under-ordering seasonal goods
- Dealing with tariff-driven cost changes
Felix’s View on LLMs in Enterprise Decision-Making
A major theme of the interview was Felix’s argument that LLMs do not belong in pricing or marketing optimization.
His reasoning
Enterprise decision automation requires systems that are:
- Deterministic
- Cheap
- Accurate
- Explainable
LLMs, in his view, do not naturally satisfy those requirements for this kind of use case.
His broader point
- Many startups are building “wrappers” around LLMs
- That may work for some use cases, but not for core enterprise optimization
- The best products will likely use LLMs as part of the stack, not as the entire solution
Important distinction
- Felix is not anti-LLM
- He believes LLMs can improve parts of software
- But he sees machine learning + explainable predictions as the right foundation for decision-making in his category
Key Lessons From the Episode
For founders
- Start with the customer problem, not the technology
- Founder-led sales can be essential for the first tranche of customers
- Early enterprise customers need patience and close communication
- A/B testing can be powerful, but it increases operational complexity
For SaaS companies
- Value-based pricing works best when you can prove impact
- Specific, high-value use cases beat broad, generic positioning
- Not every workflow should be “agentic” or LLM-driven
- Explainability matters, especially when software makes decisions for businesses
Lightning Round Highlights
Best business advice
- Use management by objectives
- Define clearly how success will be measured
Recommended book
- Master and Margarita by Bulgakov
- He likes books that help him unwind outside of work
Best money spent
- Hiring a senior backend engineer very early
Personal productivity habit
- Focus on tasks he can finish today
Passion outside work
- Beach volleyball
- He says it feels like a small vacation and sometimes even uses it as a team activity
Bottom Line
Felix’s journey shows how a deep domain background, a willingness to start with consulting, and disciplined founder-led sales can get a SaaS company from zero to $1M ARR with only 10 customers. The interview is especially useful for founders building B2B enterprise software, because it highlights:
- why data access matters early,
- how to prove value with A/B tests,
- how to price around measurable uplift,
- and why not every AI product should be built around LLMs.