Data marketplace for enterprises: every dataset behind one entry point

Large organisations rarely have too little data. What's missing is a place where you can see what exists and who's allowed to use it. So we designed and built a white-label data marketplace: one entry point where datasets can be found and access can be requested through a fixed route.

I think this is one of the most honest projects to talk about. There was no shiny AI feature and no flashy app. There was a problem almost every large organisation will recognise. This is how we tackled it.

What problem did the data marketplace have to solve?

Information was spread across platforms and storage locations, with no shared index. Staff didn't know what existed and went looking for whoever had it. Sometimes they redid work that had already been done elsewhere. Access ran through separate permissions per system, so no one had a central view of who could access what.

That last point is more than an annoyance. If no one has the overview, no one can explain why someone had access either. When an auditor or a privacy officer asks, that quickly becomes an awkward conversation.

I see this pattern more often when companies start talking about AI. Everyone wants a smart application, but first you need to know what data you have and who's allowed to use it. That's why our AI audit starts with that question too, and not with a model.

Why did we choose one entry point instead of another system?

Because yet another place to store data only makes the problem bigger. The answer was one entry point above it all: an overview of what exists, with a fixed route to get access. The data stays where it is. What changes is that you can find it and request it properly.

The platform is white-label SaaS. So it runs under the name and brand identity of the organisation using it. Users see their own environment, not a vendor's. That matters more for adoption than you might think. A tool with an unfamiliar name feels like something from outside, while a tool in your own brand feels like your own.

That choice has a commercial side too. One platform that several organisations use under their own name is a SaaS product, and that's how we set it up. Read more about our approach under SaaS development.

How did we build the data marketplace?

As a secure environment where users see which datasets exist, request access and view data without leaving the environment. Every request goes through an approval step. That records who was given access and why. The technology is a React front end on a .NET back end.

Designed on research, not assumptions

Before a single screen was drawn, we did UX research and interviews with corporates. The request process is based on that. The research decided which steps are really needed and which could go. I'm a fan of this, especially for access processes. If you design on gut feeling, you usually build one approval step too many. Then people work around it anyway.

Built by twelve specialists

Twelve team members from Your Tech Club built the platform, from the React front end to the .NET back end. The focus was on three things you can't add to a data platform afterwards: security, scalability and the ability to connect very different data sources. That last one in particular takes patience, because every source has its own quirks.

What does the design of the data marketplace look like?

Calm and businesslike, with the data in the leading role. The base is a deep purple (#6610f2) with a bright blue (#4077FD) as an accent. Every user has a personal workspace with profile details, an activity timeline and blocks showing colleagues and teams. There are also reporting screens with charts, so you don't just find data but can view it straight away. All in the customer's brand identity.

The Data reports screen, for example, shows an overview of annual revenue with blocks for orders, sales, profit and income. Another overview shows visits, clicks and sales in an area chart, with a bar chart per brand below it. So you can see on one screen what a dataset is worth, before you even open it.

We also did the branding and the accompanying website. The marketing page carries the promise 'Unleash the full power of your data', with a feature page showing the dashboard in the customer's own brand. Design and product came from one team. That saves a lot of translation work between agencies.

What does a data marketplace deliver?

One place for indexing, access management and data visualisation. Who can access what is visible centrally, instead of being scattered across systems. Staff see what exists, request it through a set process and view it in the same environment. That way, data that was already there becomes genuinely usable.

If you're torn between building and buying: Joey wrote earlier about custom software or an off-the-shelf package. And if you want to know roughly what a project like this costs, have a look at the cost of a web application. For a smaller example of how we connect scattered data, read the case about the quote tool and CRM integration for De Duurzame Jongens.