Scaling Yango Play and deciding what deserved to survive.

Context

Yango Play launched into MENA with a compelling proposition: one subscription connecting entertainment, music, and adjacent services across a wider technology ecosystem.

That breadth created its early advantage.

It also created the next problem.

As the portfolio grew, the product began accumulating competing surfaces, overlapping propositions, and features with very different usage economics. Music, video, gaming, AI recommendations, subscription management, and ecosystem benefits were all expected to coexist inside one experience.

The strategic question was no longer:

What else can we add?

It was:

Which products create durable value, which deserve independence, and which are consuming attention and capital without strengthening the portfolio?

That is a different kind of product problem.

It is not feature prioritization.

It is portfolio allocation.

The Real Problem

Yango Play did not suffer from a shortage of ideas.

It suffered from too many plausible ones.

Every surface had a reasonable internal argument:

• Music increased product breadth.
• Gaming created engagement.
• Video drove entertainment value.
• AI recommendations promised personalization.
• Subscription bundles increased perceived value.
• More navigation entry points made more content technically reachable.

But a portfolio cannot be managed through plausible arguments.

Each product needed a defensible role in the system:

• What job did it perform?
• Which behavior did it create?
• Did it improve activation, retention, ARPU, or LTV?
• Did it strengthen another product?
• Was it a feature, a standalone product, or an expensive distraction?
• What would happen if it disappeared?

The difficult part was not generating opportunities.

It was separating strategic adjacency from feature accumulation.

My Role

The broader Yango super-app organization was large and matrixed: three CPOs, multiple Heads of Product, and dedicated product managers across subscriptions, payments, money, content, growth, and core product surfaces.

I was not the sole owner of the super-app or every roadmap inside it.

My remit was more cross-cutting: new initiatives, portfolio-level discovery, product strategy, launch readiness, and post-launch evaluation across multiple products, surfaces, and markets.

I worked alongside the relevant product owners, design, analytics, engineering, content, and commercial teams to answer the questions that sat between established domains:

• Which propositions deserved investment?
• Was a concept ready to launch?
• How should a new product be positioned?
• What needed to change after launch?
• When was continued investment no longer justified?

That remit covered initiatives across Yango Play, Yango Music, Yasmina, Yango Plus, and Yango Taxi.

My role was to make uncertain product bets more defensible, not to replace the PMs accountable for their individual domains.

That meant connecting:

• Market and behavioral evidence
• Funnel and cohort analysis
• Product economics
• Interaction architecture
• Experiment results
• Technical feasibility
• Executive priorities
• Local market differences

I also managed the discovery system behind those decisions: a $100K–$300K annual budget and approximately 40 to 50 studies and experiments per year.

Research was not a separate service layer.

It was part of the portfolio operating system.

The Portfolio Thesis

The core thesis became simple:

Not every valuable experience belongs inside the same product.

Super-app logic rewards bundling.

Product logic is less sentimental.

Some behaviors benefit from aggregation. Others require a dedicated identity, habit loop, and proposition.

Music was frequent, emotional, personal, and identity-bound. It behaved like a destination product.

Gaming was episodic, weakly connected to the subscription proposition, and expensive to sustain. It behaved like an adjacency searching for a reason to exist.

Discovery was not an isolated interface problem. It was the mechanism determining whether the value already inside the portfolio could be reached.

My Vibe was not just a recommendation feature. It was a test of whether users would trust an AI system to shape something as personal as taste.

Once we framed the portfolio this way, the roadmap changed.

Decision 01 — Spin Yango Music out of the super-app

The conventional super-app answer would have been to keep music inside Yango Play.

Bundling increases apparent value. It reduces the need to acquire users twice. It allows one subscription to subsidize multiple experiences.

But our evidence showed that music-first users did not experience music as a secondary entertainment feature.

Music had a different behavioral signature:

Higher frequency
Stronger habit formation
Greater identity attachment
More language switching
More context-dependent use
More demand for personalization

People did not merely consume tracks.

They used music to regulate mood, signal identity, structure routines, and move between social contexts.

A tab inside a broader entertainment product could provide access to music.

It could not automatically create a relationship with a music product.

The product decision was to greenlight Yango Music as a standalone experience with its own acquisition funnel, onboarding, recommendation model, monetization logic, and brand relationship.

This was not a cosmetic spin-out.

It changed the unit of strategy.

Instead of asking how music increased engagement inside Yango Play, the teams could ask whether Yango Music could create its own activation, retention, and monetization loops.

The standalone product has since reached more than 60,000 active users.

The same market evidence also shaped the decision to introduce a freemium model in Egypt, where the obstacle was not simply willingness to pay.

Users needed enough time with the product to decide whether its recommendations deserved a place in their daily routine.

Decision 02 — Redesign discovery as a portfolio mechanism

As the content portfolio expanded, discoverability weakened.

The problem looked like navigation debt:

• Too many destinations
• Competing labels
• Uneven hierarchy
• Content hidden behind product boundaries
• Users defaulting to the surfaces they already knew

But navigation was only the visible layer.

The deeper issue was that the product architecture reflected the organization of the portfolio more clearly than the intent of the user.

Working with the relevant product and design teams, we redesigned the navigation model around high-frequency behaviors and clearer product entry points.

The goal was not to expose every feature equally.

It was to make the strongest reasons to return easier to reach.

That required explicit choices about hierarchy:

• Which destinations deserved persistent navigation
• Which belonged inside contextual discovery
• Which content should cross product boundaries
• Which actions represented primary intent
• Which surfaces were consuming attention without creating value

Within three months, in-app music discovery increased by 50%.

The metric mattered.

The underlying product lesson mattered more:

Discovery is not decoration around the portfolio. It is how the portfolio becomes economically real.

Content that cannot be reached does not contribute to perceived value, retention, or subscription justification.

Decision 03 — Make the AI visible enough to trust

My Vibe was the AI-native recommendation experience inside the music proposition.

The initial product instinct was to make the intelligence invisible.

The user would open the product, press play, and the model would generate a dynamic listening experience with minimal effort.

Technically, that was elegant.

Behaviorally, it was incomplete.

Early evaluation showed that users did not only judge recommendation quality.

They judged what the system appeared to believe about them.

Music taste is unusually sensitive because it is tied to:

• Identity
• Family context
• Language
• Religion
• Generation
• Mood
• Social status
• Memory

An invisible system asked users to surrender control before it had earned trust.

The solution was not simply to improve recommendation accuracy.

We redesigned the experience so users participated in constructing the model’s understanding.

The product surfaced preferences, prompted meaningful choices, and made the personalization process more legible.

This introduced deliberate friction into a category that usually treats effortlessness as an unquestioned virtue.

The wizard was not there because users could not operate the feature.

It was there because co-authorship created trust.

The product principle was:

When an AI system acts on identity, participation can be more valuable than automation.

This changed how we thought about adjacent AI experiences across the portfolio as well.

Model quality mattered, but so did confidence, control, explanation, correction, and the user’s ability to shape the outcome.

Decision 04 — Sunset gaming when the economics failed

Killing a product is usually harder than launching one.

A launch has advocates, deadlines, assets, and momentum.

A sunset has sunk cost, emotional ownership, and the fear that removing something will damage the wider proposition.

Gaming had a plausible role inside Yango Play.

It could increase session frequency, broaden entertainment value, and create lighter-weight engagement between video or music sessions.

The question was whether it actually did.

We evaluated the product through:

• Daily active usage
• Retention cohorts
• LTV contribution
• Cross-product behavior
• Subscription relevance
• Ongoing development and maintenance cost

The evidence showed limited contribution to durable engagement and weak impact on the economics of the wider portfolio.

Keeping it alive would have meant continuing to invest because the feature already existed, not because the portfolio still needed it.

I made the case to sunset gaming, grounded in usage data and observed user behavior. The decision itself sat with the CPO.

The move avoided up to $4 million in annual investment and maintenance.

More importantly, it returned product, design, engineering, and leadership attention to propositions with stronger strategic leverage.

The hardest part was not proving that gaming had low usage.

It was proving that its removal would make the portfolio stronger.

How the Decisions Were Made

The four decisions were different, but the operating model was consistent.

1. Start with the portfolio question

Every initiative had to state its role in the wider system.

Was it an acquisition engine, retention mechanism, monetization layer, ecosystem benefit, or independent product?

If a team could not state the role clearly, the roadmap was not ready.

2. Combine stated demand with revealed behavior

User interviews were useful, but preference alone did not decide investment.

We combined qualitative evidence with:

• Funnel behavior
• Cohort retention
• Usage frequency
• Cross-product movement
• Experiment results
• ARPU and LTV implications
• Market-level differences

This prevented enthusiastic feedback from being mistaken for durable demand.

3. Treat MENA as a portfolio of markets

Saudi Arabia, the UAE, Egypt, the Levant, and the wider GCC did not share one entertainment model.

Price sensitivity, household sharing, language switching, content preferences, payment behavior, and trust in recommendations varied materially.

The product strategy needed a common platform with market-specific propositions, not a fictional average regional user.

4. Separate product truth from launch momentum

A feature did not earn protection because it had shipped.

A product did not earn scale because it had executive sponsorship.

A model did not earn trust because it was technically sophisticated.

Each bet remained open to revision.

5. Make evidence operational

Insights were converted into:

• Product requirements
• Portfolio recommendations
• Experiment hypotheses
• Prioritization decisions
• Design changes
• Investment or sunset proposals

The deliverable was never the study.

The deliverable was the decision it changed.

Outcome

Yango Play grew from launch to 1.3 million subscribers and approximately 600,000 weekly active users across ten markets.

The portfolio became more deliberate:

• Yango Music moved from embedded feature to standalone product.
• Music discovery increased by 50% after the navigation redesign.
• Post-onboarding drop-off decreased by approximately 23%.
• My Vibe evolved into a more participatory AI recommendation experience.
• Egypt received a market-specific freemium model.
• Gaming was sunset, avoiding up to $4 million in annual investment.
• Product discovery became more embedded in roadmap, monetization, and portfolio decisions rather than operating as downstream validation.

I would not claim that one person or one intervention created Yango Play’s growth.

Consumer-product growth is cumulative: content, marketing, engineering, commercial partnerships, design, pricing, and product decisions all compound.

My contribution was to improve the quality of several consequential decisions inside that system, and to make sure weak assumptions were challenged before they became more expensive.

What I Would Do Differently

I would have introduced a stricter portfolio scorecard earlier.

The decisions were evidence-led, but the evidence architecture matured while the portfolio was already growing.

A stronger version from day one would have given every product and major feature an explicit score across:

Strategic role
Activation contribution
Retention contribution
ARPU or LTV impact
Cross-product leverage
Market fit
Operating cost
Opportunity cost
Confidence level
Kill criteria

That would have made three things happen sooner.

First, music’s case for independence would have become visible earlier.

Second, gaming’s weak portfolio contribution would have been harder to defend through engagement metrics alone.

Third, My Vibe’s success criteria would have included trust calibration and correction behavior from the beginning, not only recommendation consumption.

I would also have pushed freemium earlier and across more price-sensitive markets.

A recommendation product needs time to demonstrate that it understands the user.

Asking for payment before that relationship has formed reverses the natural order of trust.

The product lesson is broader than music:

Monetization should follow the moment value becomes believable, not merely the moment access begins.

Product Lessons

01  Super-app logic does not override product behavior.

02  Discovery is a portfolio and monetization mechanism.

03  AI systems acting on identity need co-authorship, not only accuracy.

04  Engagement without retention or economic contribution is not strategic value.

05  Product leadership includes reallocating attention and capital away from shipped work.

06  A portfolio becomes stronger when every product has a reason to exist, and a condition under which it should stop.

Closing Note

/* the portfolio decision */

> Add when the product creates leverage.
> Separate when the behavior demands independence.
> Redesign when the mechanism is wrong.
> Kill when the economics stop holding.

// growth is not the number of things a portfolio contains