Product Design
Growth
Research
2024 ∙ Bark.com
United Kingdom

Bark is a UK-based marketplace connecting customers with local service professionals, from photographers and personal trainers to plumbers and wedding planners. When customers submit a request, relevant professionals receive the lead and can respond with quotes. Elite Pro is Bark's premium subscription for professionals, giving them higher visibility in search rankings, priority access to leads, exclusive credit discounts, and tools like an invoice generator.

The subscription was losing subscribers. Churn was climbing and new sign-ups were declining, despite the product genuinely offering valuable features. The core issue turned out to be simpler and more fixable than anyone expected: subscribers couldn't see what they were paying for.

I led the end-to-end design for the Elite Pro relaunch, working closely with product, research, marketing, and development teams over 2 months. My job was to transform Elite Pro from a set of scattered features into a coherent, visible value proposition that subscribers could actually understand and use.

Understanding the problem

We ran two research phases: twelve qualitative interviews with current and former subscribers, and a quantitative survey with 426 participants testing feature desirability and willingness to pay. The number that stopped everyone in the room was that 61% of Elite Pro subscribers didn't know they had higher ranking in search results, one of the subscription's most valuable benefits.

Three things were broken. Benefits were invisible. The page that explained Elite Pro features was only accessible to non-subscribers, so once you paid, the information disappeared. Features were scattered across legacy locations throughout the product, making them hard to find. And after the first month, subscribers received no reminders of their Pro status or what they were getting for their money. Cancellation felt like the rational choice because nothing was making the value feel real.

Design

Once we knew the problem was visibility rather than the product itself, the direction became obvious. I didn't need to redesign Elite Pro. I needed to make it legible. The internal landing page came first: out went feature names like "Seller List Optimisation," in came plain language tested with real users. "Get seen by more customers" isn't more words, it's just honest about what the feature actually does for someone's business.

The centrepiece of the redesign was the Elite Pro dashboard — a single hub where subscribers could see all their benefits in one place with live status. Things like "5 free leads this month: 2 used, 3 remaining" made the value concrete and immediate. Subscribers also got access to an exclusive performance widget showing profile views, search hits, and lead engagements, each with a trend indicator, giving them something no free-tier professional could see: real proof that their profile was actively reaching customers. The dashboard also gave direct access to actionable features like the invoice generator and monthly free leads.

The technical constraint here was real. Rebuilding legacy flows like seller list management would have taken months of development time and disrupted existing user patterns. Rather than fight that, I designed a hub-and-spoke model. The dashboard displayed all benefits and their status, then directed users to existing legacy locations when needed. It made everything visible without requiring a full product rebuild.

I also added a Coming Soon section showing features in development. It served a specific purpose: giving hesitant subscribers a concrete reason to stay one more month rather than cancel before new capabilities arrived.

Before the new dashboard launch, I ran an unmoderated usability test with 10 Elite Pro subscribers. Participants completed tasks across both the landing page and the dashboard, testing comprehension of the benefit language and the core idea of a centralised Elite Pro hub. The results validated the direction: users understood what Elite Pro offered more quickly with the new landing page language, and the hub concept landed clearly. One thing testing flagged was that the free leads card copy was creating confusion, so we refined it before launch. It's a small detail, but the kind of thing that would have quietly undermined the dashboard's usefulness had it shipped as-is.

One more thing that reminded users they’d made the right choice was the Elite Pro monthly insights email, which gave the subscription an ongoing presence outside the product. Written in business language rather than marketing language, it showed subscribers their performance metrics, leads responded to, and ROI indicators each month. It turned a passive subscription into something that felt active and worth keeping.

When it became clear there wasn't development capacity for a custom cancellation flow, I designed the retention experience within a third-party tool's constraints. It wasn't aesthetically perfect but it was strategically sound. Launching with a basic retention flow was better than having none, and it also captured data on why people were leaving, which fed directly into future improvements.

Results

Reduction in churn within the first quarter after launch, as subscribers could finally see and access what they were paying for

More trial users converted to paid subscribers, driven by the dashboard giving them a tangible experience of Elite Pro’s value before committing

Attempted cancellations stopped by the retention flow in its first quarter, while generating actionable data on why subscribers were leaving

Reflection

I’ve worked on products where the features were genuinely weak and needed fixing. This wasn’t one of them. Elite Pro was a good product that had quietly become invisible to the people paying for it. Seeing churn drop once subscribers could simply see what they had was a strange feeling. The fix was almost embarrassingly obvious. The trial-to-paid result said the same thing differently: give people a real taste of the value and the decision makes itself.

The parts I found most useful were the constraints. The hub-and-spoke model happened because rebuilding legacy flows wasn’t on the table, not because it was the ideal architecture. Same with the third-party retention flow. Both times I would have preferred more control, but both times the shipped version generated data I wouldn’t have had otherwise. I’m more comfortable with ‘good and live’ over ‘perfect and waiting’ than I used to be.

London, UK