We celebrated the close of our sixth programme for early-stage founders (VentureCEO) with a panel event bringing together perspectives from across the ecosystem.
Edward Keelan, Partner at Octopus Ventures was joined by Fenella Boyle (exited founder & former CEO of Versed AI), Katy Wigdahl (CEO, Speechmatics), and Agata Nowicka (serial founder & Managing Partner, AI Visionaries).
The market context is clear: AI deal volumes remain steady, cheque sizes are materially larger and European AI players are setting global benchmarks. Founders are operating in an environment that changes quarterly, and the core challenge is two-fold: defensibility and go-to-market velocity.
The central question isn’t whether AI is hype or reality, it’s who, in practice, is delivering value in the new age of AI?
HYPE VS. OPPORTUNITY
We’ve moved quickly beyond the binary “bubble or boom” narrative. There is significant opportunity – especially for AI-native businesses – yet we still need to tip-toe the line between optimism and caution.
AI is becoming a hygiene factor across the economy, but the promised productivity gains aren’t uniform. Some teams see rising costs and more complexity in sales motions, for example.
And the business reality? AI wins when it solves specific, high-stakes problems deeply. Prices are compressing, engineering costs are rising, and companies need to manage squeezed margins whilst building genuine value.
Revenue can materialise rapidly in credible AI businesses – nothing like the dot-com era in terms of fundamentals – but exit pathways (especially IPOs) are scarce and buyer enthusiasm for “AI tools” is normalising.
COMPETING AS AN AI-NATIVE – OR NOT
2025 sharpened the difference between ‘AI-native’ and ‘AI-adopting’ companies. For non-native founders, there’s a playbook:
- Anchor on the customer’s workflow and ROI;
- Demonstrate value in 1-2 month cycles;
- Use your embeddedness to collect distinctive data that compounds over time;
- Think global from day one and build distributed teams capable of operating across jurisdictions.
VALUATIONS, PROOF POINTS & SECONDARIES
With limited M&A or IPO activity, clarity on “who pays next” is scarce. Founders need proof points, margin discipline and crisp, data-backed narratives. Secondary transactions help, but they don’t replace these fundamentals.
GO-TO-MARKET HAS CHANGED FOR FOUNDERS AND INVESTORS
In practice, startups aren’t selling “AI” – they’re solving concrete customer problems. What has transformed is the enterprise sales toolkit now available to startups:
- Territory planning across global prospects;
- Automated analysis of annual reports to craft messaging;
- Nuanced targeting once reserved for well-funded sales teams.
Enterprise sales now demand up to 60 interactions, where single-digit touches once sufficed. Patience, structure and precision win.
For developer-centric products? Enable quick trial testing and integration. Build presence where your audience lives. Strong social proof on platforms like G2 and Reddit matters, and sales engineers play a pivotal role in bridging credibility from first touch to production.
Investors are leveraging AI, too, but tool-based edges erode as capabilities become ubiquitous. What continues to differentiate is network, judgement, and founder support – not mass AI-generated outreach.
BUILDING DEFENSIBILITY: RELIABILITY, INSIGHT & VERTICAL DEPTH
Which moats matter?
- Production-grade reliability;
- Accuracy;
- Vertical depth.
Continuously prove your solution in contexts where the cost of failure is high and the ROI is undeniable. Many “vertical SaaS” plays, for example, look less compelling if their core function will soon be absorbed into platforms users already rely on. Ask whether your product is a feature waiting to be bundled – like spellcheck in the 1980s – or a capability that remains independent, defensible, and data-advantaged.
Founders should interrogate where agents and platform-native features will compress opportunity, and orient toward areas where proprietary data, distribution, or workflow lock-in keeps them relevant.
AGENTS VS. SAAS: FROM HYPE TO PRACTICALITY
The narrative around autonomous agents has cooled from ‘SaaS killer’ to ‘model-specific impact’. Many agent-based features fail to cross the adoption chasm, and agent hallucinations complicate trust and procurement.
In this climate, the winners solve one or two mission-critical problems end-to-end – exhibiting “radical empathy” for customer pain points. More features aren’t better – deeper outcomes are.
Startups have a superpower in responsiveness, too. They can reallocate effort faster, embed more deeply with customers, and iterate ahead of larger incumbents.
AUDIENCE INSIGHTS
GPT Wrappers? Only defensible when built on proprietary datasets with measurable outcomes. General-purpose wrappers face commoditisation.
Coding agents? They reduce grunt work, but AI engineers remain essential. Integration and connected systems deliver the real multiplier effect.
How can I raise funding fast?
- Build a clean data room (contracts, clients, ARR, retention metrics), streamline the deck.
- Be ready to respond quickly as VCs accelerate term sheet timelines. Founders with strong headline numbers control the pace.
- Have clarity on the type of investor you want – stage, sector, and support model – helps you move decisively when momentum hits.
- Working with an advisor to help you on the journey is advised.
SO… WHO’S BUILDING REAL VALUE?
- AI-native companies solving high-stakes problems with measurable outcomes.
- Products with data advantage, workflow ownership, and frictionless adoption.
- Teams with disciplined economics, global reach, and operating agility.
- Founders who use agents surgically and iterate faster than incumbents.
WANT TO TALK AI STRATEGY FOR YOUR BUSINESS?
If you would like to hear more about how we have helped multiple AI startups achieve scale through our work across tax advisory, finance support, compliance and strategic finance, get in touch.
