4 min read

Sparking connections: Engaged couples + Wedding vendors ✨

When I first joined the marketplace team, growth was slow and monetization was just starting. We shipped hundreds of impactful features and optimizations, but we needed to produce step-changing results.

Challenge

User feedback and quantitative data revealed challenges in the marketplace. Couples faced difficulty finding vendors, with 40% of couple inquiries unanswered. Vendors struggled with low incoming inquiries and lacked options to enhance their pipeline. How can we simplify vendor discovery for couples and provide vendors with more quality leads?

Couple-facing opt-in

In order to legally share couple information with vendors, we needed to determine when and how to get consent from couples. Through rounds of iteration, an explicit modal (defined by a designer on my team) was the solution that gained support from legal counsel and business stakeholders. We aimed for clarity to make the decision easy for couples.

Initially, couples would (only) opt-in after sending an inquiry to ensure they provided sufficient data, making them quality leads for vendors. To increase opt-ins and feature impact, we expanded scope to add another opt-in touchpoint: couples opting in after favoriting a vendor. However, this change presented new challenges:

1. Couples who hadn't submitted an inquiry didn't have enough data to be a quality lead, so vendors were less willing to pay for the connection.
2. With two types of couples, it was uncertain how to best present them to vendors and anticipate vendors' actions.

Vendor-facing "find couples"

Unfortunately, the orginal designer's proposals were not finding internal alignment after the scope expansion. Design time was running out, and I was hearing concerns from various directions. I assessed the situation and reached out to my partners (Maya, Nobu, Nate) to confirm that I had their support. It was time for fresh eyes and additional design muscle. I stepped in to facilitate the resourcing transition by initiating multiple conversations, most critically with the original designer and a new designer who would support my lead. The former was relieved, the latter excited, and the team was eager for next iterations.

We created initial low-fidelity mockups for internal review, then advanced to mid-fidelity for testing on usertesting.com. Each version maintained the inbox's foundational structure but introduced visual distinctions. Since favorited couples were expected to have less data, a card structure seemed appropriate. However, without actual data, it was challenging to assess scalability or predict which couple type would generate more connections.

With user input and support from engineering, we opted for a straightforward approach — using a table, with updated information architecture, sort functions, and replaced components. Conveniently, a major rebranding effort was happening simultaneously as we rolled out our new design system. The momentum from this effort allowed us to implement a fresh and functional solution with low engineering effort. We utlized new badges to signal a difference between the two couple types. These components brought a bit of color and prominence to this new information, and reduced the cognitive effort for vendors. There was no need to “reinvent the wheel." Instead, we "cleaned house" and met the requirements.

Performance

We saw encouraging usage that allowed us to continue investing in the feature:

💙 Vendors loved it!
In the first month with no marketing, 27K+ total invites were sent by 1.5k vendors
Vendor adoption grew to 22% over the next two quarters.

🔥 Connection rates for inquiries coming from the feature grew by 2x.

📈 Growth in incremental inquiries and connections continued.

From a follow-up survey, we learned that:

• 62% of couples are interested in MTFM after launch (citing less work, less vendors to sort through, chance to find vendors they might have not found, and a reduction stress of reaching out)
• 71% of vendors are using Zola more because of MTFM

[add testimonials]

Next Steps

We made multiple optimizations over the next several months. Another designer and I synthesized the initial readout from our product and business managers to iterate immediately. There were a few main drop-off points in the funnel. Our responses were a mix of straightforward best-practices in terms of product marketing, as well as bigger tests like trying a new opt-in type and adding a personalized note. For the future, we also made several hypotheses and potential solutions, with plans to test out cross-conversion tactics.