From dial-up to AI: Three decades of software development within digital signage

Today, software platforms orchestrate content, data, and experiences across thousands of screens globally. But when it all began, there was no such thing as a platform, barely even an industry.

Building in a world without infrastructure

In the late 1980s, distributing content to screens meant something very different from what it does today. When Daniel started working in the space in 1989, there was no internet to rely on. Content had to be sent via dial-up modems, connecting to each individual player one at a time.

When Fredrik entered the field a decade later, things had improved, but only slightly. Content could now be distributed via satellite or broadcast networks, typically overnight. But there was still no return channel, no feedback loop, and no guarantee that anything had actually worked.

“You sent it out and hoped for the best,” Fredrik recalls.

In this environment, software had to be built differently. Systems needed to be extremely robust, capable of operating independently without constant monitoring or intervention. Failures couldn’t be fixed instantly; they had to be anticipated and designed around.

Yet even more challenging than the technical limitations was something less obvious: no one really knew what the technology was for.

The problem was never just technical

From the outset, Dise’s ambition was clear: build software to display content on screens, and do so internationally, despite starting as a small team in Karlstad, Sweden.

But the real challenge quickly revealed itself elsewhere.

“The biggest misconception,” they explain, “was thinking this was a technical problem.”

Early projects often failed not because the technology didn’t work, but because organisations didn’t understand how to use it. Decisions were driven either by IT or by marketing, rarely by both. And crucially, there was no clear ownership of the system after deployment.

As a result, many initiatives followed the same trajectory: a strong initial idea, a successful launch, and then stagnation. No content strategy, no operational process, no long-term responsibility.

“It wasn’t about screens,” Daniel says. “It was about everything around them.”

From experimentation to maturity

The early years of digital signage were marked by experimentation and, as they describe it, a fair amount of “cowboy business.”

Concepts like retail media appeared long before the market was ready. Companies built business cases on advertising revenue that never materialised, while others underestimated the complexity of managing networks of screens.

In hindsight, the issue was often the same: a fragmented approach. Projects focused on individual components, hardware, content, or monetisation, rather than the full ecosystem required to make them work.

Over time, that changed. The industry matured, and with it came a broader understanding of what successful deployments require: cross-functional collaboration, defined processes, and a clear view of how content, technology, and operations fit together.

But even today, some of the early patterns still resurface, just in more sophisticated forms.

The quiet evolution from CMS to platform

One of the most significant shifts in the industry has been the move from simple content management systems to full-fledged platforms. Yet for Dise, this wasn’t a sudden transition.

From the beginning, their approach extended beyond managing content. It was about building a system that could create, distribute, and ultimately orchestrate experiences across screens. Over time, this naturally expanded to include data, integrations, and automation.

The key decision was not to build bespoke solutions for individual customers, but to develop a general product that could be adapted through partners. It was a principle learned through experience, custom solutions might solve immediate problems, but they rarely scale.

Instead, the focus shifted to creating a flexible core that others could build upon. What emerged was not just a product, but an ecosystem where partners play a central role in tailoring solutions to specific markets and use cases.

Integration becomes the backbone

If one development fundamentally reshaped the architecture of digital signage platforms, it was the rise of integration.

In the early days, systems were largely isolated. Content was static, and external data played a limited role. But as soon as dynamic content entered the picture, news, weather, and pricing, the need for integration became unavoidable.

What started as simple data feeds evolved into something far more complex. Today, digital signage platforms are expected to connect seamlessly with retail systems, business data, and content pipelines.

The concept of being “API-first” came later, driven not by theory but by necessity. Partners needed more control, more flexibility, and the ability to integrate the platform into broader ecosystems.

“We didn’t set out to be API-first,” Fredrik explains. “We became it because the market demanded it.”

The challenges that never disappeared

Despite decades of progress, some challenges have remained remarkably consistent.

Understanding what to show, and why, continues to be one of the hardest problems to solve. Technology can enable almost anything, but without a clear purpose, even the most advanced systems fall short.

There is also the ongoing tension between marketing and IT. Digital signage sits at the intersection of both, and success depends on aligning these perspectives from the start.

Then there is scale. Managing a handful of screens is fundamentally different from managing thousands. What works in small deployments quickly breaks down as networks grow, forcing a shift toward automation, standardisation, and integration.

And finally, there is the physical world. Unlike purely digital systems, signage operates in real environments where hardware fails, networks fluctuate, and conditions vary. This constant interaction between software and reality adds a layer of complexity that never fully disappears.

Why everyone underestimates it

One of the most consistent observations from their experience is how often newcomers underestimate the complexity of the space.

At first glance, the problem seems simple: display content on a screen. But beneath that simplicity lies a web of dependencies, content logic, scheduling, integrations, hardware constraints, and user behaviour.

“Everyone thinks it’s easy,” Fredrik says. “Until they try to build it.”

This underestimation leads to familiar mistakes: overly narrow solutions, rigid architectures, and systems that struggle to scale beyond initial deployments.

AI: a new shift, familiar patterns

With AI now entering the picture, the industry is facing another major transformation. But unlike previous waves of hype, Daniel and Fredrik approach it with cautious optimism.

AI will undoubtedly accelerate development. It reduces repetitive tasks, enables faster prototyping, and lowers the barrier to creating customised solutions.

At the same time, it introduces new challenges. Faster development increases the need for review, testing, and governance. And while AI can generate code, it doesn’t eliminate the need to understand it, especially in complex, production-critical systems.

“Developers won’t disappear,” Fredrik notes. “But they’ll need a deeper understanding of what they’re building.”

There is also a broader question of dependency. As AI tools become embedded in development workflows, they may introduce new forms of vendor lock-in and cost structures that are still not fully understood.

Building for the long term

If there is one guiding principle behind Dise’s journey, it is a reluctance to chase trends for their own sake.

Rather than being first to adopt new technologies, they have focused on adopting what proves to be useful. It is a strategy rooted in experience: building features that no one uses is simply wasted effort.

This approach has helped them navigate multiple technological shifts, from on-premise systems to SaaS, from static content to dynamic data, and now toward AI-driven workflows.

Starting over, what would change?

If they were to start from scratch today, the fundamentals would remain surprisingly similar.

The biggest difference would lie in scale. Where earlier systems were designed for hundreds or thousands of screens, today’s architectures must anticipate millions.

That requires a different mindset from the outset: cloud-native but not overly dependent, modular rather than monolithic, and designed for continuous evolution.

But the core ideas would stay the same. Build for partners. Think globally. Avoid unnecessary customisation. And always design for scale.

An industry that never stands still

After more than three decades, digital signage still resists a neat definition. It continues to evolve alongside changes in technology, retail, and user behaviour.

Perhaps that is why it has taken so long to mature and why it remains so complex.

“It never really settles,” they reflect.

And that, more than anything, may be the defining characteristic of the industry they helped build.