DIRECTV. No longer
the expensive cable bill.

The Problem

Benchmark studies had been returning the same verdict for years. DIRECTV was perceived as too expensive. Not by a little. Categorically. The brand had become shorthand for large bundles and large bills. The kind of TV package your parents had before anyone knew what Netflix was.

Between 2015 and 2023, the traditional cable bundle lost more ground every quarter. Streaming services were multiplying. Audiences were building their own viewing stacks. Sports here, prestige drama there, kids content somewhere else. They weren't unwilling to pay. They were unwilling to pay for things they didn't want.

DIRECTV had been circling the idea of skinnier bundles for a while. Genre Packs was the answer. Smaller, genre-based packages in Sports, Entertainment, Kids, News, and Spanish-language content that customers could mix and match. Each pack came bundled with streaming app subscriptions. MyEntertainment, for example, included Disney+, a Hulu bundle, and Max Basic with Ads. The idea was to make the value feel undeniable.

It would also be the first time anything like this had been introduced in the streaming space. Leadership was clear: we had to be first to market. That urgency shaped everything that came after. Including the things we'd later wish we'd had more time to get right.

My Role

I was the sole UX/UI designer on the project, leading the end-to-end transactional experience: the package selector, add-ons, confirmation flow, upsell placements, and onboarding entry points. I also overhauled the legacy purchase journey, which had accumulated years of unnecessary friction.

I was supported by a UX lead, UI lead, lead content designer, lead researcher, and our UX directors. Importantly, pricing, package structure, and naming were all still being defined while we were designing the screens. Strategy and design were evolving in parallel, which meant decisions made on Monday could be invalidated by a business call on Wednesday.

When the Rules Changed Mid-Game

The upsell scenario sounds straightforward. A user wants to watch a show that requires an add-on like STARZ. To subscribe to STARZ, they first need a genre pack. Two steps — logical if you understand the structure. Users did not.

What made it harder was that we didn't fully understand the structure either. Halfway through the project, business partnership agreements and content provider restrictions started surfacing that none of us had seen coming. Only the MySports genre pack qualified users for sports add-ons. The MiEspañol genre pack didn't qualify users for any add-on at all — they'd need an additional genre pack on top. These weren't UX decisions. They were deal structures baked in before a single screen was designed, revealed to us long after we thought we understood the product we were building.

My job became making those constraints navigable without making them feel punishing.

The First Round of Truth

The first usability test covered two distinct flows. Flow 1 is onboarding. Onboarding is designed for brand new customers signing up for Genre Packs for the very first time. Flow 2 deals with upsell. Upsell is about existing customers who already have a genre pack but want to unlock more content. Same product, two very different starting points. The results could not have been more different either.

Onboarding performed well. A bulleted list with a visible plus icon made multi-selection feel intuitive. Ease of completion scored 6.4 out of 7. Confidence hit 6.9 out of 7.

Bundled app subscriptions were the one thing that tripped people up. Half of participants weren't sure whether included apps like Disney+ were standalone downloads or tied to their device. A small confusion with a potentially big drop-off consequence.

The upsell flow told a harder story. 7 out of 8 participants struggled to complete it. Most interpreted the opening modal as the entire action and felt blindsided when paid packages appeared instead of the add-on they came for. Only 1 participant understood that a genre pack was a prerequisite. The word "Upgrade" made 6 out of 8 people hesitant. Terms like "paid lineups" and "genre pack" landed with equal confusion. Ease of completion scored 4.5 out of 7. User-friendliness scored 3.7 out of 7.

The onboarding was working. The upsell flow needed a different approach entirely.

Lost in Translation

Testing made the naming issue impossible to ignore. "Genre Packs" didn't communicate flexibility or modularity. When participants heard it, they pictured a fixed bundle, not something they could build themselves. Language around choice and customization tested significantly better. We brought a clear recommendation to stakeholders: the naming needed to change.

Then Legal weighed in. We couldn't call genre packs "packages." That word was off the table entirely for contractual reasons nobody on the design team had visibility into until it mattered. So we landed on "base packs" as the internal working term. Marketing moved forward with "Genre Packs" externally. 2 different words for the same thing, living in 2 different places.

The result is a gap that still exists today. The terminology on the TV app differs from what appears on the website. It wasn't anyone's fault specifically. It was a collision of timelines, legal constraints, and organizational decision-making happening in parallel with no single team empowered to resolve it cleanly.

Some problems don't get solved. They get managed.

When Abstract Meets Reality

The second round of testing focused entirely on the revised upsell flow. Under the direction of our UX director, we explored something bolder. Could more visual, abstract concepts help communicate the idea of building your own package? We designed explorations using different shapes, puzzle pieces, and colors to see if the metaphor could do the work that words hadn't been able to.

The working team was skeptical. We pushed to also test text-forward versions that guided users through the steps explicitly, telling them what to do and in what order. Both sets went into testing.

The abstract versions didn't land. Participants found them visually interesting but conceptually unclear. The text explorations performed better. 5 out of 6 participants completed the flow, which was an improvement. But the core confusion hadn't gone away.

The sticking point was the two-step nature of the subscription model itself. Genre Packs don't work like a traditional tiered subscription where each level includes everything below it plus more. Each genre pack is its own thing, with its own channels. They're additive, not hierarchical. You're not upgrading. You're adding. That distinction felt intuitive to us and completely foreign to most participants. They didn't understand they were subscribing to 2 separate items. They didn't understand there were 2 steps at all.

At least the next steps were clear. We needed to be literal. Tell people exactly what to do, in plain language, 1 step at a time. No metaphors. No shapes. Just words.

We also proposed breaking the flow into 2 separate modals, 1 for each step, so users would never face both decisions at once. It was the right call. However, it was deemed out of scope due to budget constraints. In the end, we had to work with what we had.

Making the Required Feel Chosen

After 2 rounds of testing, the answer wasn't a new layout or a smarter visual metaphor. It was words. Clear, direct, literal copy that told users exactly what they were looking at and exactly what to do next. Once we stopped trying to let the interface speak for itself and started writing like we were explaining it to a friend, things got noticeably clearer.

But copy alone could only do so much. Because we were stuck with a single modal — the 2-modal solution having been cut from scope — we fought for preselection as the next best thing. When a user arrived at the screen because they were trying to watch specific content, the genre pack or add-on they needed was already selected for them. They didn't have to figure out what to pick. The screen already knew why they were there and reflected it back to them. They saw the answer before they had to find it.

For users without a preselected path, warning modals surfaced only if they tried to proceed without the required selection. Sometimes the clearest design solution is the one that removes the question entirely.

What Failed

When Genre Packs launched, signups were slow. Not because the experience was broken. The content offering wasn't ready.

Leadership had made the call to get Genre Packs to market first. The idea was sound. The execution exposed some gaps that no amount of good UX could paper over. The MySports genre pack launched without local channels like NBC and CBS — 2 networks that carry a significant share of live sports. MyNews was missing ABC. MyEntertainment didn't include HBO. MiEspañol launched without Vix. MyKids didn't have Nickelodeon.

These weren't small omissions. They were the channels people specifically wanted. And when the content isn't there, the best-designed subscription flow in the world isn't going to close the deal.

The business has called it a learning opportunity about product readiness. That's fair. The gaps have since been filled, and signups have started to reflect it. A massive influx came when YouTube TV had their dispute with Disney and pulled its channels — customers needed an alternative and Genre Packs was there. The content was finally ready when the moment arrived.

Upsell numbers told a similar story. Performance was poor at launch, but it gave us something valuable — the opportunity to fix it properly. We went back to the drawing board and redesigned the entire upsell experience. Instead of sending users through a package selector and a separate add-on selector, we condensed everything into a single panel containing exactly what they needed to watch the content they came for. They could proceed straight to confirmation without having to select or remember anything. No extra steps. No extra decisions. The flow went from 5 steps down to 2. That work is currently in development and set to go live by end of Q2 2026.

What I Learned

You can design a great onboarding experience. You can fight for every word of copy, simplify a 5-step flow down to 2, test and retest until the confidence scores are where you want them. And none of it matters if the service doesn't have the content the user actually came for.

Design can lower the barrier to a decision. It cannot manufacture the reason to make one. The product has to earn that. This project taught me that good UX is only as strong as the thing it's selling.

Previous

Next