Seller numbers grew, but not success
Turning seller capability into a scalable system
// Case Study
SCOPE
Seller growth ecosystem across lifecycle stages
ROLE
Strategic direction, initiative development & cross-functional leadership
TEAMS
Product · Design · Research · Engineering · L10n
Work shown is from my tenure at Pinkoi. To respect confidentiality, visuals and product screens were recreated, and metrics have been simplified or obfuscated. All perspectives are my own and do not reflect the views of Pinkoi.
CONTEXT
By 2021, Pinkoi's seller ecosystem had grown to include tens of thousands of creative brands. Its competitive advantage came from what competitors couldn't easily replicate: a curated community offering unique, design-forward products.
These independent designers and makers were Pinkoi’s long-term moat, setting it apart from larger, commodity-focused marketplaces.
Over the years, we had built support around this community. Designer Relations teams earned sellers’ trust through meetups and workshops. My design and content teams created seller handbooks and seller feature stories.
On the surface, the ecosystem looked healthy. New sellers joined steadily, and the community felt vibrant. But underneath the growth numbers, sellers were losing confidence and quietly drifting away despite the resources and support available.
Pinkoi’s differentiation depended on sellers who could build sustainable businesses and stay. Without them, the advantage we had built was at risk.
The Challenge
What mattered to me was building a thriving seller community. Success would look different for every seller, but a healthy ecosystem meant sellers were making progress toward their goals, even at different speeds.
As we scaled, three pressure points emerged.
Manual support couldn't scale
Designer Relations teams built genuine relationships through meetups and personal guidance. But as seller numbers grew, the team’s capacity didn’t. We faced a choice between shallow support for many and deep support for a few, and neither option solved the fundamental problem.
We couldn’t see ecosystem health
We could assess individual sellers, but lacked a framework to understand the bigger picture: how sellers were distributed across lifecycle stages, which segments were struggling, and where progress was breaking down. We risked responding to visible pain points rather than underlying issues.
Feature requests kept multiplying
The wishlist kept growing, but different seller segments needed different things. Each addition served only part of the community, while the overall system became more complex.
The default path was appealing: requests were concrete, and shipping felt like progress.
But I wondered if there was a different way to think about the problem. With such a diverse seller base, would more features create clarity or just more complexity? Would we end up building solutions sellers couldn't figure out how to use?
The challenge wasn’t simply what to build. It was finding an approach that could scale while adapting to each seller’s journey.
SPIDEY SENSE
I understood sellers in a way that went beyond data.
Having been one myself, I navigated the full complexity of running a small business: online marketplaces, art shows, consignment deals, managing my own site. The uncertainty was familiar: the self-doubt, the loneliness of not knowing if you are doing it right, the invisible struggles that never made it into feedback forms.
Setting up at Renegade Craft Fair, SF (left) | Packing up after Maker Faire (right)
Years of conversations with Pinkoi sellers also taught me to read between the lines, hear what was not being said, and sense when someone was struggling even when they did not ask for help.
Watching sellers navigate the pandemic deepened this understanding. From pre-COVID stability to COVID disruption, I could see the toll: cross-border delivery disrupted, offline events canceled, buyer demand dropping. Post-COVID would bring another shift entirely. Each phase required sellers to adapt again.
I wanted to anticipate problems before they surfaced. The question that stayed with me: what could we build that would help sellers navigate not just today’s uncertainty, but whatever came next?
From hypothesis to evidence
We had data, but the patterns within it remained unread. We needed to make the invisible visible, translating observations into evidence that could build organizational conviction.
I pushed for comprehensive seller research. Our researchers conducted 12 in-depth interviews and surveyed 2,000+ sellers to understand their struggles. Our data scientists analyzed seller cohorts and progression patterns to identify where growth was stalling.
THE DISCOVERY
No single solution fits
The interviews revealed what the numbers could not.
A jewelry maker running her business part-time faced completely different challenges than an established homeware brand managing multiple channels. A new seller launching their first product needed different support than an experienced maker expanding internationally.
Struggles were universal, but each seller's path was distinct.
Tools existed. Understanding didn't.
The survey revealed a clear pattern.
Around 30% didn’t know where they stood or whether their efforts were working. Nearly a quarter struggled with marketing, either not knowing where to start or unable to improve despite trying. Over 20% felt buried in competition, invisible among thousands of products.
Sellers had access to tools but didn’t know how to use them effectively. These knowledge gaps took different forms depending on where sellers were in their journey, what kind of business they ran, and how they operated.
Four dimensions, countless combinations
The variety came from 4 factors.
Sellers differed by lifecycle stage. New sellers needed to learn how to sell. Experienced sellers needed to learn how to scale.
They differed by scale. A solo seller, a small team, and a larger operation each required different infrastructure to function effectively.
They differed by how products are made. Some sellers crafted everything from raw materials by hand. Others combined handwork with sourced components or machinery. Some designed products manufactured in factories. Each point on that spectrum came with different production realities and platform needs.
They differed by operating model. Part-time sellers juggling other commitments, full-time sellers focused on a single channel, and multi-channel brands managing complexity across platforms operated in completely different contexts.
When combined, these factors created paths so different that no single solution could serve them all.
We couldn't patch our way to scale
This explained why our previous approach wasn’t working. We had been building features for the largest pain points, prioritizing what affected the most sellers and leaving smaller segments to find workarounds.
But even within the largest segments, sellers differed too much for one-size-fits-all solutions. And the smaller segments struggling to adapt existing tools were collectively too significant to ignore.
For Pinkoi, this mattered beyond feature coverage. As a curated design platform, seller quality directly shaped platform quality. We had built our brand around unique, high-quality sellers rather than infinite supply. We needed sellers to stay and succeed, not just join and drift away.
The insight shifted our strategy. A different question emerged:
How might we help sellers help themselves at scale?
DEEPER INSIGHTS
When success stayed at the top
Understanding what sellers struggled with wasn’t enough. We needed to see where growth was breaking down, which sellers were stuck, and why they couldn’t move forward.
The research revealed an ecosystem with concentrated success and widespread struggle.
The long tail wasn’t growing
When we analyzed seller performance across revenue, reach, and traffic patterns, a clear divide emerged. Revenue was concentrated among a small group of established sellers. The majority were still building their businesses, with potential but no clear path forward.
Our growth pattern analysis went deeper, examining sellers who had joined in the previous six months. What we found:
45% had never made their first sale
33% achieved only sporadic orders, technically 'active' but losing confidence with each passing month of minimal sales
16% showed gradual improvement, receiving 1–2 orders on good days.
Only 5% showed healthy patterns: consistent orders, multiple sales per day
Where momentum broke down
Previously, I had visualized the seller lifecycle as three interconnected loops to help the team understand seller growth phases. Mapping the findings onto this framework revealed where progress was breaking down.
Many sellers got stuck in the middle retention loop. They would join, make their first few sales, but then couldn’t build momentum. When success remained inconsistent, confidence dropped, and they either became inactive or shifted focus elsewhere.
Conceptual model: Seller Growth & Retention Loops
The loops showed where progress stalled. The next question was why.
Signal lost in the noise
Analyzing seller progression revealed distinct blockers.
31% rarely logged into Shop Manager or engaged with available tools. It was a catch-22: not knowing where to start kept them away. Staying away meant they never learned what actions would make a difference.
Another 2% showed up consistently, but their products weren’t being seen. They were actively trying but couldn’t figure out how to break through the noise.
6% received decent traffic but still couldn’t convert visitors into buyers. They lacked insights to understand what was working or how to improve.
Across these stages, sellers faced the same challenge: they couldn’t determine what to prioritize. The platform provided tools, but sellers couldn’t distinguish signal from noise. They were putting in effort, but it wasn’t compounding into results.
STRATEGIC APPROACH
Getting the sequence right
Choosing capability over features
Nearly 40% of sellers had clear progression blockers, not from lack of tools, but from not knowing how to translate those tools into results. More features wouldn't solve this. Sellers needed guidance systems that could adapt to their situations and help them help themselves.
I focused the enablement team on building that capability.
This was the harder choice. It meant investing in systems that wouldn't show immediate returns, building for sellers who might never succeed or might leave regardless, and resisting internal pressure to chase visible wins.
But it was the only path that addressed the root problem. Without this foundation, the platform couldn't unlock potential across most of our seller base or sustain long-term health.
While the enablement team built infrastructure, I kept monetization selective but still delivered results. We needed revenue to follow seller success, not get ahead of it.
When we did build monetization features, the enablement infrastructure strengthened their impact. Research informed which services to prioritize. Guidance systems helped sellers use Ads and other paid services more effectively, supporting adoption and better use.
Enablement and monetization amplified each other.
SOLUTION
From tools to seller capability
The strategy centered on two types of infrastructure: enablement systems that scaled guidance across diverse seller needs, and selective services for sellers ready to invest more.
Guidance that scaled with sellers
We built three components that formed a continuous learning loop across the seller journey.
Progressive Growth Actions gave sellers structure when uncertainty was highest. New sellers faced overwhelming setup with no clear priorities. We broke the journey into milestones, from basic setup to cross-border expansion, so sellers could see where they stood and what came next.
Guidance system across dashboard and Ads workflows
What began as a lightweight advertising optimization experiment evolved into a modular automation layer spanning campaign activation prompts, bid recommendations, budget alerts, and performance diagnostics.
By 2023, the system could surface condition-based suggestions across workflows, adapting guidance to seller maturity and campaign performance.
The underlying logic remained consistent: detect meaningful signals, translate them into clear actions, and reduce hesitation at the moment of decision. The interface changed, but the behavioral model scaled across teams and use cases.
Guidance helped sellers act, but action without feedback didn’t build confidence. Once sellers tried new behaviors, they needed to know whether their efforts were working. Shop Analytics closed that loop.
Shop Analytics consolidated fragmented metrics into a single view, helping sellers develop consistent monitoring habits. Rather than refreshing everything daily, we varied update frequencies: some metrics updated hourly, others daily or weekly. This created natural check-in rhythms rather than one-time visits.
Metrics alone weren't enough. Without attribution, improvement stayed accidental rather than repeatable. We broke down traffic sources so sellers could see both their own impact and Pinkoi's contribution.
Internal sources showed how buyers found and engaged with their shops through search, browsing and interactions, AI recommendations, and featured placements. External sources showed how Pinkoi’s marketing brought new attention to shops.
Contextual guidance helped sellers understand what was working and what to adjust next.
Structure showed what to do, guidance prompted next steps, and analytics helped sellers understand the impact. Each component strengthened the others and adapted to sellers wherever they were.
This enabled us to address monetization differently.
Revenue built on readiness
Ads were already generating revenue, and there was pressure to expand paid offerings more aggressively.
But scaling monetization without a broader base of ready sellers would quickly hit a ceiling. Only a limited group could immediately benefit from advanced services. Building ahead of readiness would constrain adoption now and long-term growth.
Instead of asking what could generate revenue fastest, we asked a different question: Which seller problems were becoming solvable at scale, and which were not yet?
Early signals were emerging. More sellers were engaging frequently with Shop Manager, checking analytics, and acting on timely guidance. Selective monetization became viable.
Before building, we tested demand through surveys and in-depth interviews for three potential services: Advanced Analytics, Targeted Messages, and Custom Landing Pages.
Advanced Analytics attracted strong interest, but the free Shop Analytics baseline already met most sellers’ needs. Only a small segment needed deeper functionality, leaving a narrow addressable market and limited revenue potential. We chose not to pursue it.
Targeted Messages revealed a different pattern. Interest remained high, willingness to pay held across seller tiers, and the service aligned with behaviors sellers were beginning to adopt.
Sellers described frustrations around customer communication: no way to reach different customer groups, and no control over how coupons were distributed.
Targeted Messages addressed both. Sellers selected a goal (driving purchases, re-engaging dormant customers, or increasing awareness), and the system automatically generated audience segments and message templates, inserting buyer names, relevant products, and coupons by default.
Custom Landing Pages addressed a different execution constraint. Sellers saw opportunities for seasonal campaigns, product launches, and promotional events, but lacked the design and engineering resources to execute quickly.
Using modular templates with banners, featured products, coupons, and media, sellers could create campaign pages or tell their brand story without design or engineering support. They could launch in minutes, not weeks.
How enablement and monetization compounded
Shared infrastructure created leverage. Rather than building separate systems, we designed modular frameworks both teams could use. Guiding Tips began with Ads optimization, then extended to broader seller guidance.
Research informed both roadmaps. A survey testing interest in Targeted Messages revealed monetization priorities and seller guidance needs at the same time.
The sequence mattered. Enablement built the foundation. Monetization followed readiness, not the reverse. Enable first. Monetize second.
Results
What enablement made possible
System impact
The enablement infrastructure created a reinforcing cycle of engagement.
Sellers who engaged with one component naturally moved to the next. Those who completed Growth Actions milestones began responding to Guiding Tips. When they acted on suggestions, such as setting up coupons or adjusting listings, they checked Shop Analytics to understand the results. Then they returned for more guidance.
The pattern was visible in behavior data. Guiding Tips proved 4× more effective than push notifications, reaching sellers in context as they made decisions. Shop Analytics visits climbed 41%, and sporadic check-ins became consistent monitoring habits, with some sellers returning to Shop Manager 5–6 times daily.
For sellers just starting out, activity increased 17% in the first week. Another meaningful signal came from dormant sellers: some who had been inactive for over five months returned after viewing their analytics and began listing products again.
Seller validation
Sellers described the shift in their own words:
"Being able to see how my traffic broke down helped me understand what was actually working in my shop. It gave me a clearer picture of my overall performance."
“Most of my keywords weren't being searched at all. The suggestions showed me what to use instead, and my search ranking improved so much after that.”
"Checking Shop Manager daily and making adjustments based on what I saw started showing results after a while. That momentum made me want to keep investing effort."
"The suggestions were straightforward. I just had to decide if they made sense for my shop. I didn't have to figure everything out on my own anymore, which saved so much time."
Each seller found a different way in, but the shift was the same: less guessing, more acting on what they could actually see.
Ecosystem impact
Advanced Services launched as more sellers developed the capability to benefit from them.
Hundreds of sellers subscribed in the first month. Monthly recurring revenue reached five figures within the first quarter. Sellers who actively monitored Shop Analytics were more likely to subscribe to Advanced Services and retain their subscriptions.
These launches were part of a broader seller-services ecosystem supported by the shared enablement infrastructure. By 2023, that ecosystem had grown to around 10,000 monthly subscriptions with roughly 90% retention, contributing over 30% of platform revenue.
What this unlocked
Enablement changed the quality of growth.
Sellers who developed operational habits made better decisions. They understood what drove traffic, how conversion shifted, and when paid services could accelerate results. Monetization became an extension of competence, not a substitute for it.
By shifting from feature releases to capability building, Pinkoi evolved from supporting individual sellers to building an ecosystem where sellers and platform grew together.
// Case study
A final note
None of this would be here without Seller Service and Paid Service Squads.
Having stood on both sides, as someone who once didn't know if the effort was worth it, and as someone who got to help answer that question at scale, this one stayed with me.
