Belgian startup Dalton, which turns static websites into self-improving growth engines, today announced it has raised a €1 million Pre-Seed round from leading technology founders.
I spoke to CEO Vincent Declercq to learn more.
Websites are no longer static billboards
Dalton makes websites self-improving with AI. Its platform analyses websites, runs continuous experiments, and personalises user experiences to maximise conversion, 24/7 automatically.
Most company websites are still static billboards — unmeasured, unpersonalised, and slow to change — even as digital advertising has become highly data-driven. Giants like Amazon or Booking run thousands of experiments a year to squeeze more revenue from website visitors, but most businesses lack the people, technology, and traffic to do the same.
Dalton changes that with a specialised AI agent obsessed with increasing conversion rates. It constantly analyses a website, suggests improvements, launches experiments, and personalises experiences for each visitor segment.
A champion swimmer and a magician turned AI entrepreneurs
Dalton is led by CEO Vincent Declercq and CTO Ward Van Laer, whose unconventional backgrounds shaped their approach to entrepreneurship.
Declercq, a former elite swimmer and Belgian champion who studied Operations Research Engineering at Cornell, went on to work in venture capital at Volta Ventures and in operator roles at StuDocu and itsme, after an early start selling firecrackers as a child.
Van Laer, an AI expert and award-winning magician — you read that right — for a decade, studied Computer Science and AI at UGent and launched three side businesses, including a Shopify store during COVID that scaled to 15,000 customers.
The pair previously founded another startup that failed to find product-market fit, a learning experience they credit with shaping their resilience.
According to Declercq, the idea for Dalton actually came from Van Laer:
"After we wrapped up a previous startup that didn't work out, we sat down at the Winter Circus in Ghent and started listing things we were truly passionate about. We kept coming back to growth hacking, digital marketing, and data-driven experimentation."
Van Laer had also run three side businesses that scaled through Meta and Google ads, one of them a Shopify store. He realised he could constantly test and improve ads easily, but not his store.
"He thought, "Why isn't there something that automatically tests and optimises my website the way ads work?" With AI advancing so fast, it felt like the right moment to solve that gap. That was the spark for Dalton."
The technology behind Dalton
Dalton is a blend of generative AI, machine learning, and traditional software engineering.
According to Declercq, many AI products go too heavy on GenAI — you get a prompting box and not much else, and it doesn't always work as expected — "We wanted a real balance."
"Generative AI drives everything around suggesting and implementing website changes directly in HTML — tweaking layouts, headlines, CTAs, or adding sections."
The team has developed agent flows to ensure a smooth process. Traditional software engineering powers the UI/UX to keep it intuitive. And its machine learning algorithm handles live-traffic adjustments, which means even sites with fewer than 100,000 visitors can get meaningful results.
"That algorithm is inspired by ad optimisation and platforms like Netflix," shared Declercq.
One website, 100 versions: Dalton tailors experiences in real time
Dalton goes beyond traditional A/B testing, which tries to find the single best version of a site. Dalton takes it further with personalisation — tailoring the experience based on context.
Declercq explained that, "for example, if a user comes through a SaaS ad targeting a specific pain point, we can adapt the site so the message matches the ad. Typically, companies would need to build multiple landing pages to do this; Dalton handles it dynamically."
In practice, this might mean emphasising "vegan" or "plastic-free" for an e-commerce brand, or pushing social proof higher up on the page for a SaaS product. These micro-optimisations often deliver 10–20 per cent conversion lifts. And because attention spans are short, surfacing the right message quickly can make a huge difference.
Every change is measured, so the process is fully data-driven.
"It's like having 100 versions of your site live at once, and the system continuously learns which ones perform best."
Dalton can turn tiny tweaks into 20–40 per cent conversion gains
It cannot be underestimated how massive the impact of small changes can be. Shifting a headline or button text can deliver 20–30 per cent more conversions.
"You don't always need a whole redesign," shared Declercq.
While Dalton is still at an early stage, there's a clear pattern across customers: they spend heavily on ads.
"Whether it's e-commerce or SaaS, they're paying for traffic but not seeing proportional results," shared Declercq.
"Our sweet spot is mid-sized companies — think product-led growth SaaS firms like Aikido where users sign up, request demos, or fill out forms.
But we also work with very small businesses, though changes take longer, and very large enterprises with full testing teams."
For the latter, Dalton helps them go from running a handful of tests per month to ten times that volume, unlocking AI-driven scale in their workflows.
One of Dalton's customers increased conversions by 40 per cent in just weeks - fully automated.
Dalton's human-in-the-loop AI approach
However, Dalton's approach isn't just about agentic AI.
"Our philosophy is that Dalton should do 99 per cent of the work, while users approve final changes before they go live," said Declercq.
"That keeps control and minimises risk. The system acts like an employee: you provide your site, it generates changes, and you approve them.
Once enough data is gathered, it tells you what worked, what didn't, and proposes the next set of optimisations."
I was interested in how Dalton ensures continuity. For example, what if I regularly shop on the same website and want a consistent experience?
If you're on the same device, Dalton drops a cookie so you see the same version you saw last time. While the continuity can't be guaranteed across different devices, most changes are subtle — like adding a badge that says "vegan" — so users often don't even notice.
The real challenge, however, is not technical but cultural.
"Traditional A/B testing is finite — you start an experiment, stop it, and implement or discard the result," explains Declercq.
"Dalton is continuous. There's no 'end.' That requires a shift in how growth teams think about website optimisation."
The web is ripe for evolution with personalised adaptability
Declercq contends that websites are one of the oldest digital products, but fundamentally, they haven't evolved much:
"The future, as we see it, is fluid, personalised sites that adapt dynamically to each visitor — like an ad manager, but for websites.
Some argue websites will disappear into chat interfaces. I don't believe that. Companies will always need a storefront — a place where customers can browse and explore beyond a single conversation."
The company also sees opportunities around accessibility and insights. Regulations are coming — possibly a GDPR-style moment — requiring sites to meet accessibility standards. Dalton could make compliance seamless. Similarly, instead of overwhelming users with dashboards, Dalton can surface direct insights:
"20 per cent of mobile users in this country are having a poor experience. That shift from raw data to actionable guidance is where the real value lies."
There's also scope around accessibility, which is often overlooked, but regulation is on the horizon. For Dalton, it would be a natural extension — flagging issues and fixing them automatically.
There are also adjacent opportunities, such as detecting "rage clicks" when users get stuck, or identifying non-clickable elements that frustrate users.
The company also sees Dalton evolving into a kind of growth analytics agent — replacing clunky dashboards with clear insights. Instead of digging through Google Analytics, Dalton tells you what's broken and why.
"Dalton is applying AI to a real, overlooked problem that can drive meaningful growth for businesses.
We're excited to keep building and to see Dalton become the go-to solution for website optimisation."
Investors in the current round include Syndicate One and a roster of prominent entrepreneurs, such as:
- Stijn Christiaens (Collibra),
- Matthias Geeroms (Lighthouse),
- Jorn Vanysacker (Henchman),
- Gilles Mattelin (Henchman),
- Wouter Van Respaille (Henchman),
- Joris De Wit (Teamleader),
- Roeland Delrue (Aikido),
- Louis Jonckheere (Wintercircus),
- Bruno Vandegehuchte (Wintercircus),
- Alexander Stevens (Greenomy),
- Christophe Morbee (NewSchool VC), and e-commerce leaders.
Robin Wauters, founding member of Syndicate One, added:
"Dalton demonstrates a well-defined problem, impressive early traction, and a highly capable, fast-acting team. These qualities are exactly what investors look for in an early-stage company.
Additionally, the widespread support from prominent founders across Belgium indicates strong external validation and confidence in Dalton's potential."
Louis Jonckheere, founder of Wintercircus and backer of Dalton, said:
"Where Dalton sits at the intersection of AI, marketing, and product is exactly the kind of frontier we look for in Wintercircus. These founders are building smart and delivering tangible results for real customers."
With customers including Sunday, Dexxter, Aikido and klimaatshop.nl, Dalton is expanding quickly. Backed by the new investment, the team will focus on product development from its base at Wintercircus Ghent, while adding engineering and product talent to keep pace with growth.
Would you like to write the first comment?
Login to post comments