Paloma is the leading venture studio in ANZ. Our 100+ team has forged an AUD $250M+ Venture Portfolio, and helped steer firms like Afterpay to billion-dollar acquisitions.
There’s a few topics that come up repeatedly in our conversations with early stage founders when they’re figuring out how to build product. In November we (Seb Cox and Simon Plant) hosted a session at Blackbird Sunrise Aotearoa to tackle a few of these themes head on.
For those who missed the live session, here are our views on some of the common challenges founders face when building startups:
1. MVP Is Dead, Long Live MVP
Building a successful product is one of the hardest challenges for any founder. The question of what to take to market can be overwhelming. Too often, outdated “lean startup” thinking leads founders astray:
Go too lean, and your MVP (minimum viable product) may fail to solve any real problem for customers. The market dismisses your idea when, in reality, was it the test itself that was flawed?
Go too heavy, and you risk locking in assumptions that prove wrong, wasting precious resources and limiting runway for course correction.
The sweet spot lies in preparation. Founders who succeed know their market deeply, develop thoughtful hypotheses, and assess the risks upfront rather than optimising for experimentation.
Think Risk, Not Features
When deciding what to build, shift your focus from features to risks. Three key risks to evaluate are:
Technology/Product Risk – Can you build what you need with the resources you have?
Market Risk – Will customers want and pay for it?
Business Model Risk – Is there a solid business model under your product? Are the unit economics good? How much does it cost to acquire customers?
Understanding these risks highlights gaps, helps prioritise resources, and identifies your key differentiators. The takeaway? Do the prep work. Deeply understand your customers, risks, and business model. When it’s time to launch, you’ll move faster and reduce the chance of building the wrong thing.
2. Listen to Customer Feedback or Trust Your Gut?
It’s common advice to build what customers tell you. We think this is wrong. Everyone wants to make data driven decisions but in most cases startups do not have enough data to do this approach properly.
Gathering enough statistically significant data takes time to collect
Building products by consensus or trying to please everyone dilutes focus and impact.
Great products are born from outlier insights, not crowd consensus. Henry Ford’s famous “faster horses” quote reminds us of the risk of listening too literally. Instead, founders should focus on identifying patterns, deeply understanding their customers' underlying needs & having the conviction to set a direction.
Sometimes, what you don’t build is just as important as what you do.
3. Beware the Commoditisation of AI
AI tools, especially LLMs (large language models), are exciting enablers, but they come with risks. If your product relies heavily on third-party AI remember that everyone has the same tools that you do, which means that thousands of teams are building similar AI-powered products.
Low barriers to entry mean fierce competition.
‘Arms race’ type strategies are fine but you need to find ways to maintain your marketshare if you manage to achieve it through the speed and versatility of AI. You also need to be aware of incumbents punching down into your vertical. They have the same access to the tools and a lot more data. You validating an opportunity & them stepping into the space is always a risk.
There is even a world where the companies that own the models begin to specialise & will cannibalise many existing businesses.
Remember: AI isn’t a business model; it’s a tool. To succeed, build a strong foundation that doesn’t depend solely on AI’s novelty. Compete with unique value and a robust go-to-market plan.
4. Humans Are Often Cheaper Than Code
Technology is expensive, but building the wrong technology is even costlier. Founders often over-invest in automation and scalability too early, solving problems they don’t yet have.
Instead, embrace “Flintstoning” — using scrappy, manual solutions behind a veneer & enabled by technology. This hands-on approach helps founders:
Understand customer behaviour and product usage firsthand.
Avoid premature investments in automation that may not align with customer needs.
Scaling too early is a risk; solving today’s problems first is the smarter play.
5. Low-Code: A Powerful Start, but not always a Long-Term Solution
Low-code and no-code tools like Bubble, Airtable, Webflow, or Framer have revolutionised early-stage building:
They are great for idea validation, prototyping, static sites, tightly scoped tooling however;
Challenges arise with scaling, collaboration becomes a bottleneck quicker than you think and security & performance often suffer.
Adopt a low-code mindset: empower non-technical team members to test and iterate quickly, build in a modular way, so business operations aren’t disrupted when re-platforming or intensive customisation needs to happen. Most importantly; know when it’s time to re-platform for growth.
6. Tech Stack: Simple Is Smart
Your tech stack matters, but not as much as you think. Early-stage startups should prioritise:
Mainstream tools, languages & frameworks – Leverage established technologies with strong community support to save time and hiring headaches.
Simplicity – Avoid overly complex solutions that often indicate misplaced priorities.
Cloud credits wisely – Use startup credits but be mindful of vendor lock-in.
Ultimately, productivity and flexibility trump chasing the latest tools.
7. LLMs: An Enabler, Not a Silver Bullet
LLMs are transforming productivity and enabling new products, but they’re not magic. Founders should:
Avoid over-reliance on a single AI vendor to mitigate lock-in risks.
Understand the limitations of LLMs (e.g., hallucinations, self-confirming loops).
Combine AI with experienced engineering to maximize results.
Understand the cost implications of training & tuning on their underlying COGS - the sands under LLMs are constantly shifting & you will need to keep up to compete.
While LLMs can boost efficiency, the cost of chaining operations or handling errors can eat into margins, particularly for low-margin businesses.
8. Optimizing for Mindset and Experience
In early-stage teams, mindset is just as important as technical experience. Look for people who:
Are flexible and focused on outputs - people who are too rigid or idealogical about a specific framework or way of working often struggle in early stage startups.
Balance early-career energy with seasoned expertise.
Prioritise ruthlessly and make tough calls about what not to build.
Great startups don’t over-engineer or chase complexity. They solve real problems with elegant solutions—and that starts with a strong team mindset on output.
