SMART TALK

AI Edition

We sit down with GDR Chief Data Officer Ulrik Larsen to discuss the thinking behind IDfree.ai.

How is AI transforming audience activation?

Why may the future of marketing be less about tracking individuals and more about understanding patterns.

Ulrik Larsen

Chief Data Officer, GDR

Ulrik Larsen is the driving force behind GDR's technology development team.

Co-founder and Chief Data Officer Ulrik Larsen is one of the key minds behind GDR's technology, data strategy, and product innovation.

A specialist in programmatic advertising, statistical modelling, adtech, and audience intelligence, Ulrik leads the teams responsible for turning complex data challenges into practical solutions for marketers.

Known for his compliance-first mindset and passion for simplifying workflows, Ulrik plays a central role in developing GDR's privacy-first technologies, including IDfree.ai.

Before we begin, let's start with the big picture.

AI is dominating conversations across the marketing industry.

However, beneath the headlines, many advertisers and agencies are still searching for practical ways to work smarter, move faster, and deliver better results.

Ulrik took time out for us to answer

"Why GDR decided that now is the right time to for IDfree.ai."

And what problem it is designed to solve.

Everyone seems to be talking about AI right now. Why launch IDfree.ai?

We're launching IDfree.ai because marketers don't need more complexity. They need less.

The industry we're part of has been adding layers for years: more tools, platforms, and IDs; new workflows; integrations everywhere; and, on top of that, new processes for most of us.

At the same time, agencies and media buyers are under pressure to move faster, prove results, and navigate a fragmented ecosystem.

We looked at that reality and asked ourselves:

"What if custom audience activation could be as simple as describing the audience you want to reach? What if the client could do it? What is the youngest programmatic rookie could do it?"

That's what useful AI does. It takes tasks that used to require a specialist with technical know-how and makes them accessible to anyone.

That's the thinking behind IDfree.ai.

What problem are you solving?

Countless times we've seen how thoughtful and strategic briefs from advertisers or client directors get stripped down and butchered at the programmatic trading desk.

Not out of bad will but from lack of time, and the need for specialist knowledge to translate a brief into a coherent targeting strategy.

Now, you just ask idfree.ai in plain language what you want.

"Coffee lovers who visit cafes weekly and appreciate specialty brews" almost always become "like coffee" in the trading desk.

We set out to do three things in idfree.ai:

1. Keep the complexity from the briefs.

2. Lower the data costs.

And finally

3. Make sure it never takes more than 60 seconds end-to-end.

That is the three core principles for the dev team:

Smarter, cheaper and much faster.

Any idea that did not support all three was thrown away.

How does GDR fit into this story?

In a word, GDR is the reason IDfree.ai exists.

It's at GDR that we've spent years building audience intelligence across markets and lifestyles, behaviours, and of course, geographies; our geo-intelligence supercharges the right reach.

We've never relied on cookies, mobile IDs, or personal identifiers in our approach to solutions.

Instead, we've focused on understanding patterns.

That's why one of our core messages is:

"We See Patterns. Not People."

IDfree.ai isn't a departure from GDR.

It's the next step.

It's built on the same principles, the same audience intelligence, and the same privacy-first methodology GDR has been based on from day 1.

Many marketers feel overwhelmed by the pace of change. What would you say to them?

I'm not minimising anyone's challenges, but I'd say that this is actually a great moment to be in marketing.

Yes, things are changing quickly. But we also have the opportunity to rethink old beliefs.

For a long time, our industry was hooked on tracking individuals as a shortcut to what we called "insight" - and overlooked the patterns that were telling more relevant stories.

But today, marketers can lean on much more durable and privacy-friendly approaches - and truly scalable ones too.

That's exciting.

To us, technology is a pretty good tool to help us adapt to change.

So AI shouldn't be seen as a threat?

Not at all. AI can now process long and complicated workflows without the strange hallucinations we saw in the early days.

When used correctly, AI tools turn very demanding processes that could take days into a secure path that takes less than 60 seconds, like inside IDfree.ai.

The one thing AI cannot replace is knowledge and understanding.

You need to understand what you are doing - and especially "what good looks like". That is now the core competencies, not coding or Excel skills.

The expertise still matters.

The strategy still matters.

And the human understanding for sure still matters.

What makes IDfree.ai different?

Two things come to mind immediately: automation and data layers.

The first comes from many, many conversations with customers and users.

"Just make it easy", was a line we heard many times. We tried to eliminate all manual and bureaucratic steps.

The data deal is piped into your DSP of choice, you run the campaign, costs are baked in, no invoices and no surprises.

The second is the new 6-layer data model that now incorporates neighbourhood fingerprints, Satellite Intelligence, and Street Intelligence.

The 46 m. parameter model for the Nordics is a leap in quality over what we have seen in geographical intelligence.

If I could add a third, it would be ease of access.

There exist many models and many complicated workflows.

Ours is simple. It only requires a simple, plain language input.

No statistical or technical skills or expertise.

Can you tell us how it works?

Sure. You log in, pick a market - say Norway - and describe your audience in plain language.

Something like "affluent families in suburban Oslo interested in outdoor activities."

Hit send, about 30 seconds later, you get a complete audience report with:

1. A demographic breakdown.

2. A lifestyle profile.

3. Representative images.

4. A word cloud of what defines this audience.

5. A heat map showing exactly where they are - down to the postal code.

If something isn't right, you tell the AI what to adjust. "More urban." or "Exclude students." It re-scores instantly.

When you're happy, you set your reach, pick a DSP - like The Trade Desk, DV360, Adform - and activate.

You get a Deal ID, and your trader can go live with it immediately.

The whole thing, from typing a brief to having an activation-ready deal, takes under a minute.

And building audiences is completely free - you only pay when you choose to activate.

We didn't want to build a tool that requires training or onboarding.

If you can write a sentence, you can use IDfree.ai.

What does it cost to try IDfree.ai?

It costs nothing. And I mean that literally.

You can build as many audiences as you want, explore every market, edit and re-score as many times as you like - all for free.

There is no cost until you decide to activate an audience to a DSP.

When you do activate, the pricing is transparent:

- €0.40 flat CPM or 9% of media spend -

You decide which model works for your campaign.

And the invoicing happens automatically inside your DSP.

No separate invoice, no surprises.

We designed it this way because we want people to explore without hesitation.

The best way to understand what IDfree.ai can do is to try it. Not read about it.

Which platforms does it work with?

IDfree.ai works with all the major platforms.

The Trade Desk, DV360, Adform, Amazon DSP, you name it.

We are connected to more than 85 international DSPs - and we're expanding.

Meta integrations are on the roadmap.

What are early users saying?

The reaction that keeps coming back is "I didn't expect it to be this fast."

Planners who are used to spending days translating a brief into targetable segments are doing it in under a minute.

One team told us they built audiences for four Nordic markets in 15 minutes - work that would normally take days of back-and-forth with data teams.

What surprises people most isn't the AI part. It's the simplicity.

They expect complexity and instead get something that just works.

We're in a soft launch right now, onboarding advertisers and agencies by invitation.

If you're seeing this and want to try it, please reach out to gunnar.kihl@globaldataresources.io.

We'll get you set up the same day.

What should marketers expect next from IDfree.ai?

A few things I can share.

1. We're building a brief intake module that lets you upload an actual client brief - a PowerPoint, a PDF, a strategy doc - and the platform interprets it and builds audiences from it automatically.

So instead of translating a brief into targeting parameters yourself, you just drop the file in and let the AI read it. That's coming soon.

2. We're expanding beyond the Nordics. We have audience data and scoring infrastructure for markets like Germany, Austria, and Switzerland in the pipeline.

The architecture is built to be market-agnostic, so adding a new country is a data question, not a technology question.

3. This one excites me personally! We're connecting IDfree.ai to social media activation. Today, we activate through programmatic DSPs. But the same audience intelligence should work just as well for Meta, and we're building that bridge.

And then there's the bigger picture.

We're in the final phases of building and testing an IDfree MCP server - that's a standard that lets AI agents talk to platforms directly.

I think by the end of 2026, we'll see agent-to-agent campaign management become real.

An AI planning assistant at an agency will be able to talk to IDfree.ai, build multiple audiences, and activate the deals - without a human clicking through screens.

That's the direction the industry is heading, and we want to be ready for it.

So yes! We're just getting started.

But it's not abstract.

These are concrete things we're building right now.

Where does AI in advertising go from here?

I think we're about to see a shift from AI as a feature to AI as an operating layer.

Right now, most platforms have added AI as a button somewhere - "generate copy," "optimise bids," "suggest audiences."

That's useful, but it's still humans driving every step.

What's coming is different. Standards like MCP - the Model Context Protocol - are making it possible for AI agents to interact with platforms directly.

Not through a dashboard, but through a structured conversation between systems.

We're building for that future.

Our MCP server will let an AI assistant - whether it's Claude, ChatGPT, Copilot, or Adform's FLOW agent - connect to IDfree.ai and do everything a human user can do: interpret a brief, build an audience, review the results, adjust, and activate.

The marketers are still in charge.

They set the strategy, and they approve the activation.

But the manual steps between "I know who I want to reach" and "the campaign is live" - those can be handled by agents.

I think by late 2026, early 2027, this will go from experimental to standard.

And the platforms that are ready for it will have a real advantage.

We intend to be one of them.

One final thought?

AI rewards knowledge and insights. It punishes routine work and manual processes.

It is also incredible at minimising noise in the workflow.

The old ways of handing over to a planner who again hands over to an activation team will be gone.

This is not doom for marketing and tech.

AI will just massively reward the advertisers and agencies that know what they want and recognise "what good looks like".

Luckily, our customers and users know exactly that.

Where to begin?

If any of this resonates, go to idfree.com.

Read about how the model works, and apply for early access from Gunnar: gunnar.kihl@globaldataresources.io.

We'd love to hear from you.