Key findings in 60 seconds
- The metric that matters after launch is feature velocity: how long a capability takes to move from internal approval to a subscriber's screen.
- Two questions measure it — deployment frequency and lead time. You can ask both this afternoon.
- On a white-label platform, your release schedule belongs to the vendor. On a custom build, it belongs to you.
- Every device you add widens the gap between shared-component and separate-codebase architectures.
- The broadcaster who launched fastest often becomes the one shipping slowest — the velocity inversion.
6 weeks on the calendar. 8 weeks on the quote.
That was the whole problem, and it fitted into two numbers. A broadcaster I work with needed to let viewers pick between two commentary languages on the same live stream, the kind of change that is a configuration decision on a modern stack, measured in days. Their white-label vendor came back with 8 weeks. The rights window for that broadcast opened in six.
So the match went out in one language. The audience it had been built for watched it on a competitor’s app instead, and by the time the capability was technically available, the fixture that justified it had already been played.
Nobody in that story did anything careless. The request was reasonable, the vendor’s estimate was probably honest, and the tech director had approved it the moment it landed on his desk. He simply found out — at the worst possible moment — that the clock was never his.
“Someone inside the broadcaster says yes, and then discovers how long yes actually takes when the timeline belongs to somebody else.”
I have watched that same shape repeat across dozens of projects. The details are never the same twice. Offline downloads on one device type. A betting integration racing a rights window. A payment method nobody thought about until the app was live in a new market. But the moment is identical every time: someone inside the broadcaster says yes, and then discovers how long yes actually takes when the timeline belongs to somebody else.
Why launch speed is the wrong metric for your OTT platform
Because you already launched.
The industry still talks about “time to market” as though it means launch speed — how fast an app reaches 5 stores across every device. By that measure white-label wins outright, and the evidence is hard to argue with: 6 to 12 weeks from contract to live, branded customisation included.
But most European streaming operators went live 2 to 4 years ago. That race is finished, and nobody is scoring it any more. The number that decides things now is how long it takes to move a capability from internal approval to a subscriber’s screen, across every platform you serve. Call it feature velocity. Almost nobody measures it until the week they urgently need something and find out they can’t have it.
How to measure feature velocity on a streaming platform
2 questions, and you can ask both this afternoon.
How often does your platform deploy? The 2025 DORA report, based on responses from nearly 5,000 technology professionals, found roughly 23% of teams now deploy at least once a day. If your honest answer is “whenever the vendor pushes an update,” you are receiving changes monthly or quarterly. That pace is set by the architecture underneath you.
What is your lead time? From internal approval to a subscriber using the thing, how many weeks? If the path runs through another company’s scoping, planning, QA and release calendar, you are counting in months. When you own the pipeline, a change approved on Monday is in sprint planning by Wednesday and in testing within a fortnight.
Who controls your OTT platform release schedule?
On a white-label platform, not you — and Friday night is when you feel it.
Store submissions and device certification run on the vendor’s schedule, which is fine until subscribers start hitting crashes during a live match. You cannot push the fix yourself. You raise a ticket. And while you wait for someone else’s release window to open, so does your audience, in real time, on the one night of the week you cannot afford it.
“You raise a ticket. And while you wait for someone else’s release window to open, so does your audience.”
Why does every device you add widen the gap?
Because on separate codebases, every feature ships as many times as you have platforms.
A typical broadcaster serves iOS, Android, web, Fire TV, Roku, Samsung Tizen and LG webOS, often Apple TV and Android TV as well. 7 to 10 targets. If those are maintained separately, a capability arrives one platform at a time: iOS first, Android next, connected TV whenever there is capacity. On a modular architecture with a shared component layer, the change ships once and lands everywhere. Every device you add makes that difference larger.
What is the velocity inversion in OTT platform development?
It is the moment the broadcaster who launched fastest becomes the one shipping slowest.
This is the part that catches tech directors off guard. White-label puts you live in weeks. 18 months later, every change you want routes through the vendor’s backlog and arrives on the vendor’s release schedule, so you move at their pace indefinitely. Meanwhile the team that built its own platform spent 12 to 18 months getting to production — and then started controlling its own clock.
Early on, the trade is obviously worth taking. You are live and your competitor is still in development. A year on, your competitor is live too, shipping weekly, and you are waiting for Q3. Fora Soft reached a similar conclusion in their 2026 build-vs-buy playbook: for hybrid monetisation across SVOD, AVOD and FAST, a custom build pays for itself within 18 to 24 months.
When is white-label still the right answer?
When the catalogue is straightforward and the product doesn’t need to differentiate.
I would rather say that plainly than pretend the alternative is free. A custom build takes 12 to 18 months to reach production, and for a number of broadcasters white-label remains the correct decision on the numbers.
But if you are competing for the same subscribers as Netflix, which deploys continuously, or DAZN, which is spending north of $3 billion on live rights in 2026 and runs its own front end, the question sharpens considerably. Can you afford to answer your market next year rather than this season?
Score your feature velocity position
The one number to start tracking
Feature velocity comes down to a single question of ownership: if you control the pipeline, you ship when the market needs it. If you don’t, you ship when the vendor reaches you.
The broadcaster with the 2 commentary languages measures it now. He watched a six-week window close against an eight-week quote and never wanted to be told that number again.
If your honest answer to “how long would this take?” is “let me ask the vendor,” the timeline already belongs to someone else. That is a conversation worth having at IBC this year, Stand 5.F51.
Frequently asked questions
How do I calculate feature velocity for my streaming platform?
Track two numbers: deployment frequency (how often code reaches subscribers) and lead time (calendar days from internal approval to a feature going live on all platforms). The 2025 DORA report benchmarks elite teams at multiple deploys per day. If your answer depends on when the vendor next schedules a release, you are measuring their velocity, not yours.
Can I improve feature velocity on a white-label OTT platform?
Within limits. You can streamline your own approval process and reduce the back-and-forth on specs. But the release pipeline, store submissions, device certification and QA cycle belong to the vendor. Those are structural constraints, not process problems. The ceiling is their architecture, and no amount of internal efficiency changes that.
When does a custom-built OTT platform become faster than white-label?
Typically 12 to 18 months after production launch. Before that, white-label is faster to market by definition. After that, the custom build controls its own release cycle and starts compounding the advantage with every sprint. Fora Soft's 2026 analysis puts the financial break-even at 18 to 24 months for hybrid monetisation models.
