We’re just over a week into 2026, and it’s hard to ignore how much has shifted over the last 12 months. The direction of travel is clear: more outputs, more platforms, tighter budgets, and higher expectations – all at the same time. The organisations that will be strongest this year will be the ones that can scale up and down cleanly, integrate faster with partners, and keep a tight grip on unit cost – without lowering the bar on quality.

That’s the backdrop – and it’s why reflecting on 2025 matters. Last year was when a few “future” conversations became normal operations.

Throughout 2025, one theme kept coming up in customer conversations: how do you stand out in a congested market? The competition isn’t just other broadcasters anymore. It’s social platforms, independent creators, and an endless supply of always-on content.

That pressure landed on top of a reality the industry has been living for a while: advertising and subscriber revenues remain under strain, investment is harder to justify, and many organisations are continuing to restructure – cost take-outs, new operating models, splits, mergers and consolidation. That trend hasn’t gone away as we enter 2026.

So, what did 2025 really show us?

1) Efficiency became the job, not the project

For years we’ve talked about “doing more with less.” In 2025, it became day-to-day decision making.

Operational cost is shaping production choices, staffing models and procurement. The question I got most wasn’t “what’s next?” – it was how do we save money and stay flexible without compromising output?

Cost control can feel like a threat – but it can also be the opportunity to redesign workflows around today’s realities: fewer bespoke builds, faster deployment, and systems that don’t need rebuilding every time the format mix changes.

2) Hardware still matters – but it’s rarely standalone

Hardware isn’t going anywhere, especially for large-scale international events. Live sport continues to deliver audiences and revenue, and it’s still one of the few true appointment-to-view categories.

What changed in 2025 is how hardware is used. Tier-1 productions are still grounded in FPGA hardware infrastructure – major events produced in trucks with high-density processing and switchers such as K-Frame – but more of those productions are then extended using software to create additional outputs and revenue-friendly variants – all while reducing footprint and cost.

Regional and opt-out feeds. Live branding and regionalisation. Format conversion and distribution packaging. Supplementary shows, shoulder programming and digital-first feeds.

This is where platforms like AMPP show up in practical ways – not necessarily replacing the truck, but taking on the extra production, processing and versioning work so teams can produce more from the same event with less additional kit, less friction, and a smaller footprint.

And we’re seeing the same outside sport too. Major venues and enterprise production teams are pushing for broadcast-grade consistency because audience expectations are set by the best. One enterprise client reflected recently that the production quality of their earnings calls can directly impact their share price. Quality isn’t a debate – it’s a requirement.

3) Software-defined crossed the line from “possible” to “proven”

Software-defined production has been building quietly for years. In 2025, it crossed a threshold.

I spoke at the SVG Summit in December about this: the debate is no longer whether software-defined can support tier-1 work. It can. The real question is implementation: on-prem, in the cloud, or a blend – and the answer is usually “it depends”: the event, the constraints, the risk profile, and the cost model.

The best deployments I’ve seen aren’t ideological. They start with outcomes, then work back from real requirements: latency, resilience, staffing, cost predictability, and how quickly you can scale without redesigning the whole system.

A strong proof point this year was seeing end-to-end software-defined projects such as DMC Production with Ziggo Sport, recognised not just for innovation, but for measurable impact – including footprint reduction and sustainability.

And at scale, the signal is even clearer: AMPP usage grew to an average 18 million hours usage per month by the end of 2025. That doesn’t happen because software-defined is trendy – it happens because it’s being used in real operations, for real workloads, under real pressure – with operational and business upside.

4) Sports and news stayed resilient – but the rest of the business still needs answers

Advertising confidence still concentrates around live sports and news. They’ve carried a lot of the load while other categories fragment. But sports and news can’t shoulder the business alone.

In 2025, more organisations pushed harder on new formats with lower-friction production models, faster turnaround of content, and better reuse of assets and metadata to extend the lifetime value of what they already create. Multi-rendition distribution is now a default, not a special project.

That’s also why playout and content management are being pulled into strategic conversations. When everything is expected everywhere, all at once, these aren’t background systems – they influence speed, monetisation and operational efficiency.

5) The winning approach is openness

As the drive toward lower cost of operation and faster ROI continues, we’re seeing more demand for single-vendor or limited-vendor solutions. Historically that comes at the expense of best-of-breed.

My view is that the better answer is a strong platform model: reduce complexity and integration effort, while keeping the freedom to choose the best tools.

That’s how we think about AMPP – a common platform that brings operational consistency and shared compute, while still enabling best-of-breed choices from Grass Valley and alliance partners – all while supporting standard IO and integration at the edge with your existing systems – no different to how you deploy today.

What I’m watching for 2026

1) Hybrid becomes the default build pattern

In 2026, “hybrid” should stop being a headline. Modern systems will be designed as a mix of dedicated hardware, COTS, private compute and cloud – operated as one model.

The winners won’t be the ones forcing a single approach. They’ll be the ones that let teams choose the best execution path per production, while keeping operations consistent.

2) Openness accelerates – JT-DMF and MXL will make integration less painful

The industry is aligning around a shared need: faster integration, less reinvention, and more freedom to combine best-in-class capability.

That’s why we’re investing directly in JT-DMF and MXL, supported by the European Broadcasting Union (EBU), AMWA and the Linux Foundation. These aren’t theoretical exercises – they’re practical mechanisms for making multi-vendor workflows easier to build and easier to maintain.

AMPP has been designed to align with that direction, mapping to the JT-DMF reference architecture: open, secure APIs and an architecture built to integrate. As JT-DMF recommendations evolve, the prize is straightforward: more interoperable building blocks and less integration drag.

3) Compute costs will force FinOps into production operations

Compute is becoming more contested – not just because production is scaling, but because AI is competing for the same resources – ultimately driving the widely forecasted cost increases for 2026.

Whether you rely on COTS appliances, hyperscalers, or a mix, cost visibility and control will matter more. FinOps has already been moving from finance into operations as we’re hearing directly from members of our GVX advisory committee. This is becoming an operational competency, not an optional initiative.

That reinforces requirements like sharing compute efficiently between productions, scaling without permanently carrying peak capacity, and making cost-to-produce visible at the workflow level.

This is also where platforms like AMPP help in practical terms: pooling resources across events, scaling when needed, and keeping control of the operating model without trading off output quality.

4) AI shifts from “demo mode” to “solve the right problems”

2025 was full of AI demos – useful, sometimes noisy.

In 2026, the tone is more grounded. Broadcasters are clearer on what they want AI to do: reduce repetitive work, speed up workflows, get more value from metadata, and automate tasks that don’t require human judgement – but only where the ROI is clear, because AI itself isn’t free.

More and more, the ask isn’t “build me a feature.” It’s: “give me a platform I can integrate into – so we can apply AI (and process) where it makes sense for us.”

That’s where I expect progress – and it’s consistent with how AMPP is built: secure openness, well-governed APIs, and architectures that allow customers to introduce AI agents and internal tooling to orchestrate workflows without turning the production chain into an experiment.

Closing thought

If 2025 was the year the industry accepted that old cost structures don’t survive new revenue realities, then 2026 is the year our clients turn that into an advantage.

The organisations that win won’t abandon quality. They’ll protect it – while modernising how they produce it: hybrid by default, open by design, and disciplined about cost.

If you’re seeing similar shifts (or different ones), I’d love to hear your thoughts.