Overview
A fan profile that survives interruption: onboarding that resumes two steps in, not eleven, built once and worn by two brands, DAZN and Kayo.
- Role
- UX Designer, user experience and engagement
- Company
- DAZN, with Kayo
- Platforms
- Web, mobile and TV
- Scope
- Fan onboarding, profile management and the engagement features built on top
- Onboarding
- Personalization
- Profiles
- Cross-brand design
The brief
Most onboarding resets to zero the minute you leave. This one picks up two steps in, not eleven.
I worked as a UX Designer on user experience and engagement: fan onboarding, profile management, and the engagement features built on top of them, for both DAZN and Kayo.
DAZN does not really have an onboarding screen. It has a fan profile that follows someone from the moment they subscribe through everything they do afterward: which teams they follow, what gets recommended to them, and how Kayo, a separate DAZN-owned product, treats that same person on a different app. This case study covers the onboarding itself and the profile layer underneath it, on which everything else described here is built.
The FIFA World Cup brings a wave of new subscribers who have never used DAZN. They are easy to sign up. Before the product can deliver any of its personalization promises, someone has to tell it who they are: which teams, which interests, how they want to be spoken to. That is what this onboarding captures, and it is also the moment when most products get impatient and skip it, or turn it into a chore.
The default instinct with onboarding is to make it shorter. I went in a different direction: keep it multi-step, but make it interruption-proof. A fan who closes the tab after the avatar screen and returns the next day should not have to make all the decisions again.
How five platforms personalize
Personalization is the new normal. What differs is how each platform gets there.
I looked at how five other platforms approach personalization and fan engagement, using public reporting and product documentation, not anything internal. I wanted a clear view of where DAZN's approach actually sits before deciding what to build on top of it.
DAZN
Conversational personalization lets fans take control of their viewing. Fans can request highlights, select interests and interact with live features. It is immediate, not passive, and not based only on past behavior.
Netflix
Builds personalization through ongoing behavioral learning. View history, watch time and engagement signals refine recommendations for each profile over time.
ESPN
Tailors to fan identification. Fans follow their favorite teams, leagues and competitions to get the news, scores and highlights that matter without endless scrolling.
JioHotstar
AI-assisted discovery and conversational search extend personalization, combining intelligent search with stated preferences to reach relevant content faster.
Prime Video
Focuses on contextual discovery during live events: real-time insights, match information and overlays that do not interrupt the live sports experience.
AI clip tools, an adjacent market
AI-powered media platforms are shifting discovery from browsing to conversation. Natural language search and automated highlight creation reduce navigation effort and get people to meaningful moments faster.
Five platforms, one open question: who does the customizing?
Compared side by side, the differences are not really about which platform has better technology. They are about who does the personalizing: the platform, watching quietly over time, or the fan, saying something outright.
| DAZN | Netflix | ESPN | JioHotstar | Prime Video | |
|---|---|---|---|---|---|
| Personalization | Preference-led onboarding with tailored content discovery | Behavioral recommendations | Team and league interest personalization | Interest-based recommendations | Behavioral recommendations |
| Fan engagement | Polls, predictions and interactive match experiences | Continue Watching, My List, viewing reminders | Scores, highlights, fantasy, fan content | Live chat, polls and cricket engagement | X-Ray, live insights and watch features |
| Profile model | Individual fan profiles with preference memory | Independent household viewing profiles | Account-based personalization | Multi-profile support for households | Household profiles with separate recommendations |
| Discovery | Editorial curation with personalized live content | Algorithmic recommendation rows | Team-first sports hubs | Personalized rails and Continue Watching | Recommendation rails and live sports hubs |
Comparison of DAZN, Netflix, ESPN, JioHotstar and Prime Video by personalization, fan engagement, profile model and discovery.
Most of these platforms personalize passively, over time, from viewing behavior. DAZN asks about fan preferences early in onboarding, before there is any viewing history to learn from. That is the opportunity this case study is about: strengthening the connection between onboarding, stated preferences and content discovery, so personalization compounds across every session instead of only starting once there is enough history to work with.
Findings and principles
Four findings, taken from the shipped screens themselves
This was not based on user interviews or internal analytics. I did not have access to either on this project. What follows is a direct read of interface patterns, interaction flows and observed behavior in the shipped product, so it is a design read of the interface, not primary research.
Resume onboarding picks up where you left off if you are interrupted during account setup.
The flow makes a clear commitment to session continuity.
Steps remaining should always be reflected in progress indicators. Not everywhere yet.
Profile components follow a consistent pattern across the onboarding experience.
That points to a reusable, scalable profile framework instead of one-off screens.
A shared component strategy increases consistency and reduces maintenance across future experiences.
Generated clips become part of the viewing history and are kept after they are created.
Personalization extends beyond recommendations to fan-specific content on demand.
The content the fan asked for is the content the fan already had: same history, same access.
Subscription value is introduced through interactive fan experiences, not a separate destination.
Engagement and premium content access are linked, not parallel.
Participation is part of the premium viewing experience, at the moment it counts.
Four principles, decided once, not screen by screen
Resumed, not restarted
Progress is saved when a session is interrupted, so a fan can resume onboarding without repeating a single completed step.
Build once, use everywhere
A reusable profile foundation for predictable experiences, DAZN and Kayo included, that cuts duplicate design and engineering effort.
Make, don't just suggest
Fan intent becomes custom content. A request becomes a meaningful viewing moment, not just a recommendation.
Engagement is part of the premium experience
Interactive features add value to a subscription because participation becomes part of live football instead of sitting next to it as a separate feature.
One profile, two brands
A single profile, a structured journey, two brands
Everything is driven by a common profile architecture. When fans onboard, they choose an avatar, favorite teams and personal preferences. These become a reusable profile layer that supports profile management and personalization across DAZN and Kayo, while each keeps its own look and feel.
- Account leads to Fan onboarding, which leads to the Profile system.
- Fan onboarding covers the avatar, teams, age and interests.
- The Profile system covers who is watching, and managing, editing, deleting and favoriting profiles.
- DAZN and Kayo run on the same engine underneath. The branding moves. The structure stays.
Five stages, and the one in the middle where most products would have you start again
This case study lives in stage three. It is fine for a session to end early. What most onboarding misses is what comes next.
- Successful payment. The entry point, with no personalization yet.
- Personalize. Avatar, teams, demographics and mood.
- Interruption. The session ends early. This is where most products start over.
- Restarted. The flow continues with the remaining steps and does not repeat what is already done.
- Fan profile finished. A home screen customized from the first load.
Seven flows, and what each one is for
- Fan profile onboarding: Subscription, Profile, Preferences, Customize profile.
- Onboarding resume: Interrupted session, Continue progress, Finish onboarding.
- FIFA+ interstitials: Welcome, Continue in app or web, Enter experience.
- Managing profiles: Favorites, Edit preferences, Manage profiles, Who is watching.
- Fan vote: Select a player, Submit a prediction, See results.
- Content discovery: Prompt, AI highlight, Watch, Saved history.
- Error and recovery: Validation error, Inline help, Successful completion.
Fan profile onboarding
Four screens, and nothing unnecessary on them
A new subscriber with no preferences gets a generic first session, however good the recommendation engine underneath is. Onboarding asks only for what it needs to customize that first session: an avatar, teams and a couple of light preference signals. It stays lean enough that skipping it never feels like the easier option.
Resuming onboarding
The moment I was planning for was not the onboarding. It was the disruption in the middle of it.
Personalization is almost never completed in a single sitting. A fan closes a tab, a call comes in, or the match they were about to check has already started. If onboarding has no memory of where they were, the fastest way through is to skip it altogether, and everything they had already done is lost, along with whatever the product had learned.
By default, multi-step onboarding is fragile. A returning fan has to repeat every completed step from the beginning to reach the personalized experience that was promised in the first place. Saving progress meant designing for the interrupted state as the normal case, not an edge case to tolerate afterward.
Laying out every resume state side by side did something I had not planned for: it surfaced inconsistencies in the shipped flow itself, places where the interface had drifted slightly from one resume point to another.
That is the first principle in action: resume sessions, do not restart them. It took building enough resume states to test that claim before it became clear where the product was not yet delivering on it.
FIFA+ inside DAZN
FIFA+ had to feel like an extension of DAZN, not an app to go looking for
Content from FIFA+ risked feeling like a destination fans had to go to, rather than something that was naturally part of DAZN. This journey explores several TV entry experiences that bring FIFA+ inside DAZN while also guiding new viewers to the fastest real path to watching. The fastest path is a little different depending on which screen someone opened it on.
Profiles across two brands
A household product takes more than one profile. This one had to run on two brands at once.
A household streaming product has to support several viewers on one account, each with their own profile, preferences and favorites. This brief went beyond DAZN. Kayo, a separate DAZN product, needed the same profile foundation without losing its own brand on screen.
This is the second principle, literally: build the profile system once and scale it everywhere. The alternative, a second profile system built from scratch for Kayo, would have taken real design and engineering time to produce something fans would have had to relearn anyway.
MultiAngle voting
Watching a match is passive by nature. Voting on who the camera follows is not.
In traditional live streaming, a fan has no real control over what they watch or how they experience it. MultiAngle takes a small viewing decision and turns it into something a fan takes part in. It also moves the subscription conversation to the point where the feature actually delivers value: after the vote, not instead of it.
AI highlight clips
Fans can only search for highlights they already know about. This lets them describe one instead.
Manual browsing, filtering and searching is the traditional way to find highlights, and it only ever finds moments a fan already knew to look for. This flips the orientation: describe a moment in natural language and get a personalized highlight generated on demand, without a menu or category in the way.
Problem
Fans have to scroll, filter and search manually for traditional highlights. They can only find moments they already know where to look for.
Design objective
Instead of browsing categories and menus, fans describe what they want in natural language and get a personalized highlight generated on demand.
Errors and recovery
Never let a profile fail without a reason
Validation has one job here: to explain the problem immediately and in place, without making anyone guess what went wrong or lose their spot in the flow.
One interaction language
The part that has to hold up everywhere at once
In this case study each flow investigates a different product problem. None of them invented a new way to confirm an action, load, or recover from an error. They all reuse the same underlying patterns, which is a big part of why a fan arriving through onboarding, MultiAngle or the AI clip generator still sees the same product every time.
Delete confirmation
Every critical action is contextual from start to finish. Confirmation, success and recovery are one continuous experience, not separate screens.
Responsive behavior
The same interaction model adapts across mobile, desktop and TV, respecting the strengths of each platform instead of reinventing the experience for each surface.
Loading states
Every decision point has explicit loading, validation, success and error states, giving people continuous confidence about what the system is doing.
One profile and avatar system, two brands
Avatar selection, profile management, favorites and inline validation all share the same component structure beneath DAZN's and Kayo's separate visual identities. That is what lets two products that look nothing alike deliver a consistent experience under one system of hierarchy, feedback, progressive disclosure and confirmation, wherever a fan enters it.
What I can and cannot support
I would rather keep the distinction between what I can support and what I cannot yet than let confident language blur it. This case study is based on examining the shipped product, not research I ran myself.
I do not have a measured outcome for any of the seven flows above. There is no lift in completion rate for resume onboarding, no engagement number for voting on MultiAngle, and no adoption figure for the AI clip generator.
What I can say is what each was designed to do: preserve progress instead of discarding it, turn a passive viewing moment into an active one, and let a fan create content on demand instead of only being fed what the algorithm had already decided. A single profile, created once, has carried a fan through onboarding, both brands and all the engagement features built on top of it since. The point is not that any one screen is the product. It is that they all remember the same thing: this fan was personalized once.
Personalized once. Recalled every time after.