Hey Builders, we're back!

How to raise your next round by being hard to copy: Raluca Ragab and Andrei Danescu on the five-year test

When AI can turn a good idea into working software in a matter of weeks, being hard to copy is what gets you funded. Here’s what Eurazeo’s Raluca Ragab and Dexory’s Andrei Danescu shared at How to Web Conference 2026 about the five-year test, plus five swaps to make before your next raise.

Four or five years ago, Andrei Danescu only needed one word to send investors heading for the exit. Whenever he mentioned robots, he told the How to Web Conference audience, investors would “physically run away”. In October 2025, the company he co-founded, Dexory, closed a $165m round (100m in equity and $65m in debt) to build out its autonomous warehouse robots, which scan facilities overnight so teams know exactly where everything is by the morning shift.

The round was led by Eurazeo, and the person who led it for Eurazeo joined him on stage. Raluca Ragab is a Partner at Eurazeo Growth, where the team typically leads rounds of $20m to $75m. Together, they talked through what got Dexory funded, what investors are looking for now and what it takes to scale once the money is in. 

Why the investor rulebook changed

For most of the last decade, Raluca explained, tech investors were trained to say no to hardware. Asset-light software was the dream, because if it takes off, the economics scale in a way physical products rarely do. What changed that thinking is AI – as tools like vibe coding have made software dramatically faster and cheaper to build, a lot of software-only companies now find themselves in what she called a red ocean, where anyone with a good idea can produce something very similar in a matter of weeks.

Her conclusion is that a company now needs one of two things to be defensible: some form of deeper technology that anchors its value over time, or a go-to-market so strong that its relationships with customers become the moat. Very few software companies manage the second, which is why, in her words, investors have had to unlearn a lot of what they thought they knew and look again at businesses where the laws of physics still apply, expertise takes years to build and capital itself starts to act as a barrier.

The five-year test

Underneath all of this sits one question Raluca asks of every company: will it still be around, and still winning, in five to ten years? The reason is refreshingly practical – an investor might be convinced a company is a winner today, but they will need to sell their stake to somebody else about five years later, and that buyer has to believe there are another five to ten years of success ahead.

For founders, it reframes the pitch. The question is less about how impressive the product looks today and more about why it will be hard for anyone to catch up.

How Dexory earned it

Dexory started building what we now call physical AI around ten years ago, long before the term caught on. On stage, Andrei walked through some of the decisions that helped the company grow.

It sold speed to value, from day one – Dexory sells into large enterprises that are cautious and constantly pitched by other vendors. Rather than asking them to commit millions before seeing any return, it promised customers they would see value within a week or two in a market where automation projects usually take around six months just to get running. Raluca’s take was that quick, visible ROI is exactly what makes a product fast to adopt and sticky afterwards, because customers see the benefit every day.

It refused to do free pilots – Raluca pointed to an MIT study suggesting that 95% of pilots fail and Andrei’s answer was to avoid the free pilot altogether. His reasoning? If a customer can’t pay for the product now, they won’t be able to afford it at scale and anything given away tends to be seen as lower value. Instead, Dexory agrees upfront which KPIs would make the ROI work on a three-year contract and builds the pilot around them, so a successful pilot leads naturally into a long-term deal. It also stays very close to customers during that period, because that’s where the real learning happens.

It built only what made it unique – Dexory designs, engineers and manufactures most of its product in house, but it started out buying off-the-shelf components and only switched to building its own once that stopped getting it to the next level of performance and unit economics. Buying isn’t always the answer, but building everything yourself isn’t the silver bullet either.

What scaling can teach

Today Dexory is growing roughly threefold year on year at scale, with deployments from the UK to Romania, hundreds of robots in the US and projects in Australia and the Middle East, all without offices in most of those markets. A few lessons from getting there stood out:

The first was actually a mistake. Early on, Dexory leaned towards hiring people with deep industry experience and Andrei now believes it should have focused more on people who are hungry and genuinely believe in the change the product brings, rather than those attached to the way things have always been done.

The second was that a new category needs educating. Because customers didn’t yet know what the product could do, marketing had to work hard from the start and the most valuable stage turned out to be industry trade shows. Alongside that, Dexory made case studies, ROI analysis and permission to talk about each project part of every customer relationship, which gave its sales team the references enterprise buyers always ask for.

The third was the US. Andrei pushed back on the idea that you need to spend a fortune to break into it, but described the market as ruthless about product-market fit and incredibly fast to adopt. Dexory’s first sale there, to GE Appliances, happened with no office and no local team. When Andrei asked if they wanted to see the product first, the answer was that they believed it would work, and if it didn’t, they simply wouldn’t pay. As Raluca summed it up, going to the US early can be the best reality check a product will get.

The takeaway: swap this for that

Put the investor’s view and the founder’s experience side by side, and a few swaps stand out for anyone preparing to raise:

A free pilot → a paid pilot built around the results that would justify a three-year contract.

“Trust us, the value will come” → value a customer can see within a week or two.

A product that’s easy to copy → deeper tech or a go-to-market nobody can match.

Industry veterans by default → hungry people who believe in what you’re building.

A pitch about today → a case for why you’ll still be winning in five to ten years.

Each one is something Dexory actually did on the way to a nine-figure round, and something the investor who led that round says she looks for.

Get the next How to _ in your inbox. One theme, one standout talk and one thing you can use, every week. Subscribe to our newsletter.

Sign up to receive the latest updates and a custom discount code on tickets

You may also like

How to grow your product by being the odd one out: Dave Harland’s interrupt, intrigue, convert framework

When AI can turn a good idea into working software in a matter of weeks, being hard to copy is what gets you funded. Here’s what Eurazeo’s Raluca Ragab and Dexory’s Andrei Danescu shared at How to Web Conference 2026 about the five-year test, plus five swaps to make before your next raise. Four or… Read more »

Read more

The How to Web Conference 2026 partner ecosystem: capital, tech & scale on the floor

When AI can turn a good idea into working software in a matter of weeks, being hard to copy is what gets you funded. Here’s what Eurazeo’s Raluca Ragab and Dexory’s Andrei Danescu shared at How to Web Conference 2026 about the five-year test, plus five swaps to make before your next raise. Four or… Read more »

Read more