Build Fast & Build Right
Thanks to everyone who signed up to our networking event. We hope everyone had a good time, met some new people and learnt something from our amazing panel. Here we have summarised the key discussion points and lessons…
Artificial intelligence is no longer the realm of experiments in research labs. The startup ecosystem has rapidly matured, shifting its focus from the models themselves to the way those models are applied, deployed, and scaled in the real world. At our recent panel discussion brought together experts to explore where startups should focus their energy, from the latest AI trends, to what “AI readiness” really means, to practical strategies for scaling with discipline.
What Trends Are We Seeing In The AI Space?
What was once a race to build ever-bigger models is giving way to a new era where the real competition lies in applications. Startups are increasingly focusing on last-mile, domain-specific solutions, from healthcare and retail to sales operations. This evolution is forcing young companies to mature faster than ever with the ecosystem as a whole has growing more sophisticated too. Founders, investors, and customers developing a better understanding of cost structures and what distinguishes a truly great AI business from a traditional SaaS play.
Still, the competitive landscape is far from settled. AI value creation is emerging from three fronts: nimble startups, established incumbents, and the foundational model giants. While companies like OpenAI and Google have made waves in areas like coding assistants and image generation, their impact outside of those domains remains limited. “There’s a wide open field where the giants haven’t yet dominated,” Tom observed. “That’s the space startups can claim if they move quickly and smartly.”
For Candice, the strategic challenge is clear: every startup must ask whether it is in the “firing line” of big tech or carving out a defensible niche. In her view, data has become the core battleground. “Data is no longer a by-product of your business. it is the business,” she said. Those who establish early feedback loops with customers and bake trust and privacy into their products from the beginning will have the best chance of building a lasting advantage.
The throughline across these perspectives is unmistakable: AI startups are at an inflection point. The winners will be those who shift their focus from building models to building meaningful use cases, who combine speed with responsibility, and who treat data and trust not as afterthoughts but as foundational assets.
What Does “AI Readiness” Really Mean?
“AI readiness” is one of those phrases that gets thrown around in pitch decks and boardrooms, but what does it actually look like when put into practice? For startups building in today’s AI ecosystem, it is not just about plugging into the latest model, it is about rethinking the very foundations of how a company operates.
At the most practical level, readiness starts with making AI part of everyday work. From drafting code to summarising research, the most effective teams are already using AI to generate “first drafts.” As Tom put it, “If you’re not using AI to accelerate the basics, you’re already falling behind.” But readiness also means realism: AI often tops out at 80% accuracy. Human oversight is not optional, it is the layer that ensures quality and trust.
For new companies. Culture may be the hardest piece of the puzzle. AI readiness is not just a technical capability, but a mindset shift. If you have the opportunity to start from scratch then you can embrace automation, experiment with AI-first processes, and see intelligent agents as the default rather than the exception are laying the groundwork for an “agent-first” tomorrow. This is of course, much more difficult for established companies who need to rework processes from the ground up.
The next challenge is structural. Companies that treat data as an afterthought will find themselves at a disadvantage, while those that prioritise clean pipelines and proprietary data assets from day one will have a lasting edge. Candice remarked that “In the past most of our clients were second time founders who understood that data could accelerate growth, today that’s nearly every founder. Data adoption is coming much earlier in company development than ever before.” In her eyes, AI readiness is inseparable from a company’s ability to generate and refine insights directly from its customers.
How to Scale Smarter?
Much of the conversation around AI has focused on speed, but Tom stressed the importance of slowing down in order to go fast later. He described how the Fixer team dedicated several days to cementing key business definitions, ensuring everyone was aligned and “singing from the same hymn sheet.” By taking this time upfront, the team avoided confusion down the line. For Tom, success is not just about growing fast it is about growing right. The companies that thrive are those that balance ambition with discipline, pairing cutting-edge tools with thoughtful execution.
Manal built on this point, reminding the audience that while frameworks and technology are critical, lived experience is just as valuable. The lessons learned from other founders; their missteps, surprises, and hard-won wins, can save precious time and resources. Scaling smarter, she argued, is not a formula to follow but a mindset to adopt. Building a strong network around you, she added, can provide both perspective and practical guidance during moments of uncertainty.
Anne highlighted that success often comes after several cycles of trying and failing, emphasising the need to test, tweak, and embrace trial and error when searching for the right solution. Programs like Google for Startups not only provide free credits but also enterprise-grade tooling, giving young companies the chance to experiment with AI while minimising risk. These resources, Anne noted, can make a real difference for teams navigating the complexity of scaling in an uncertain environment.
When and How to Think About Data
For AI startups, the question is not whether data matters, it is when and how to approach it.
Tom and Candice emphasised that “thinking about data” doesn’t mean over-engineering from day one. Instead, they recommended starting small with what Candice called “minimum viable data.” This approach means capturing only the most essential signals at first, enough to build, test, and learn quickly, and layering in complexity over time. Trying to design the perfect system too early, Tom warned, can waste energy and slow a team down. The goal is to let data evolve alongside the business, rather than constrain it.
Manal closed the discussion by reflecting on the investor perspective. To her, the strongest founders are those who strike the balance between ambition and pragmatism: they recognise data as their future differentiator, but they resist the temptation to boil the ocean too soon. “What matters most,” she said, “is not just having data, but learning from it, turning raw signals into feedback loops that guide smarter decisions.”
In the end, the message was clear: AI startups must treat data as central from day one, but they must also be disciplined about how they grow their data capabilities. Start small, build on solid infrastructure, and let the system scale with the business. It is not about having all the answers immediately It is about laying the groundwork for insights to emerge as the company scales.
How 173tech Can Help
173tech is a London-based analytics agency that helps growing businesses worldwide leverage data. Our core team previously led analytics at dating giant Bumble and we have since been using data to accelerate growth for the likes of MeUndies, Fyxer AI, World Of Peppa Pig, MUBI, Octopus Legacy and many more.
We typically work with fast-growing companies who want to ensure that in their speed they stay sustainable. We can help you to:
Define the right data strategy
Get In Touch
Our friendly team are always on hand to answer questions, troubleshoot problems and point you in the right direction.
