Software can be copied. A compounding data moat can’t.

There’s a question almost every AI startup is going to have to answer at some point: What happens when someone bigger decides to do what you’re doing?

Or, as someone once bluntly put it: “Won’t the LLMs just eat your lunch??”

It’s a fair question. AI is making software faster and cheaper to build. Features that once took months to develop can be replicated in a few clicks. And if you’ve built something great that works, there’s a good chance somebody else can – and will – try to replicate it.

So if your entire competitive advantage is the software you’ve built, you potentially have a problem. Building good software isn’t enough anymore. You need to build something that becomes harder to replicate over time.

And that’s where the data comes in.

Picture a castle. Now picture the moat around it.

The castle is the product. It’s what everyone can see: the features, interface, workflows and tools that customers interact with. But the thing protecting the castle is the moat around it.

In our case, that moat is data. Not scraped internet content, not generic information that anybody can access. Real operational data generated through the day-to-day activity happening across a commerce platform.

Products, suppliers, orders, materials, compliance, customer behaviour, returns, purchasing decisions. Every interaction adds another piece of information. And, crucially, those pieces don’t exist in isolation. As they accumulate and connect, the dataset becomes more useful.

This is where things start to get interesting…

More data. Better software. More customers. More impact. Repeat.

If a brand uses a platform to order products, the system starts building a picture of those products.Their materials, suppliers, environmental impact, compliance requirements and commercial performance.

Add another brand and there’s more information. Then another. And another.

The dataset gets deeper and richer, and the software has more context to work with. It can identify patterns, make connections and provide better information back to the brands using it.

The software is what brings customers in. Their use of it creates data. That data makes the software more intelligent and more valuable. And every time that cycle repeats, the moat gets a little wider.

Brands can use that information to answer questions about what to buy, how much to buy and who from. And because the data is live, there’s still time to act on it instead of pulling together disconnected information after the orders have already been placed.

Having this data before the order is placed means they save time, reduce costly mistakes, manage compliance and make better commercial decisions. That makes the software very valuable. 

And this is where the compounding starts.

More brands create more data.
More data makes the software smarter.
Smarter software creates more value for brands.
More value attracts more brands.

And repeat.

The advantage doesn’t stand still. Every new customer, product, supplier interaction and workflow has the potential to strengthen the intelligence layer underneath the software.

So what makes it defensible?

Software features can be copied. Interfaces can be recreated. AI models will continue to become more capable.

But years of proprietary operational data are much harder to reproduce. An LLM can’t invent a history of real products, suppliers, purchasing behaviour, workflows and outcomes. That information has to come from somewhere.

And when it’s generated through the actual workflows brands use to run their businesses, the value isn’t just in the volume of data, it’s in the connections between it.

That’s the moat.

The software is what brings customers in. Their use of it creates data. That data makes the software more intelligent and more valuable. And every time that cycle repeats, the moat gets a little wider.

What does that look like in practice?

Our current pilot with the European Outdoor Group (EOG) is a good example of how that process can begin.

Retail and ecommerce brands deal with huge amounts of product and supply-chain data. It arrives from different suppliers, in different formats and across different systems. Through the EOG pilot, vvast takes that messy information and standardises it into a common structure using a hard fought existing standard ‘SDEX’.

Brands can then use the data while they’re building orders, seeing information such as plastic and carbon impact before an order is placed rather than months later when somebody needs to produce a report.

For the individual brand, the immediate value is obvious – less manual work and better information at the point of decision. But something else is happening underneath.

Every product, material, supplier and order adds to the structured operational data flowing through the platform. As more brands use it, that creates the potential for a richer picture of how products are sourced, ordered and managed – intelligence that can feed back into the platform and make it more useful again.

The pilot solves a very real problem today while beginning to build something much bigger for tomorrow. 

There are a limited number of places available for outdoor ecommerce and retails brands to get involved. There’s no cost but a whole lot of value to be gained. We’ve estimated we can save brands 800+ hours of manual compliance and data processing per season, while also giving them the information to make sharper ordering decisions with the sustainability impact of each choice visible before the order is placed.

Learn more and sign up here.

T