Why Building the Store Dashboard Took 85% of Development Time

When building a consumer marketplace, the natural instinct is to perfect the consumer experience. Build a beautiful app, add social features, implement sophisticated recommendation algorithms, create delightful micro-interactions. But for Buen Provecho, nearly 90% of their development resources from 2023 through 2024 went toward building tools for stores, not consumers.

On the Levels Podcast, Guillermo Martinez, COO and co-founder of Buen Provecho, explained why they made this counterintuitive choice—and why it was the right call for solving their fundamental supply constraint problem.

The Resource Allocation Decision

Most consumer apps dedicate the overwhelming majority of their development effort to the user-facing product. The consumer experience drives downloads, engagement, retention, and revenue. Improving the consumer app creates compounding returns through better conversion rates and viral growth.

Buen Provecho flipped this conventional wisdom.

"For example, I'd say almost all 2023 and most of 2024, we dedicated 85, 90-ish percent of our dev time resources to the app for the source."

Think about what this means. Nearly two full years where almost every feature request from consumers got pushed to the backlog. Nearly two years where the engineering team focused primarily on building web applications and dashboards that consumers would never see. Nearly two years betting that solving the supply side would matter more than perfecting the demand side.

This wasn't a decision made lightly or early. They started with a functional consumer app that worked well enough to prove the concept. They had validated demand through their WhatsApp community and initial app launch. The question was: where should marginal development effort go to maximize growth?

The answer wasn't obvious. They could have focused on consumer features like personalization, social sharing, gamification, or improved discovery. These would likely have increased engagement metrics and possibly improved retention somewhat. But none of them would solve the fundamental problem: not enough stores publishing enough products.

What Stores Actually Needed

The decision to focus on store tools came from spending extensive time talking to stores and understanding their pain points. Guillermo emphasized how much this shaped their product roadmap.

"As I spend a lot of time talking to clients, it's like, need this, I want that, like add this for me. Obviously, they are not the best advisors, but they tell you the problem. So you find them a solution."

This is sophisticated product thinking. Stores would request specific features, but the team's job was understanding the underlying problem those feature requests represented. A store asking for "easier photo uploads" might actually be frustrated with the entire product publishing workflow. A request for "better analytics" might reveal confusion about how the platform was performing for them.

The problems stores faced fell into several categories. Publishing products needed to be faster and simpler, since store staff were busy and wouldn't spend ten minutes uploading each item. Inventory management had to work seamlessly, allowing stores to update quantities and mark items sold without confusion. Analytics needed to clearly show stores how much revenue they were recovering and how much waste they were preventing, making the value proposition tangible. The onboarding process required streamlining so new stores could get up and running quickly without extensive training. And integration with existing point-of-sale systems would reduce friction for chains and franchises operating multiple locations.

Each of these problems, if solved, would remove barriers to stores using the platform more frequently and publishing more products. That directly addressed the supply constraint that limited growth.

The Compound Effect of Supply-Side Investment

Improving the store experience creates compounding returns that aren't immediately obvious. When a store finds the platform easy to use, they publish products more frequently. More frequent publishing means more products available throughout the day rather than single time windows. More available products means better consumer experiences and improved retention metrics. Better retention means the consumer base grows organically through word-of-mouth without burning users through bad experiences.

This creates a flywheel. Better store tools lead to more engaged stores, which leads to more supply, which improves consumer experience, which drives consumer growth, which makes the platform more valuable to stores, which makes store acquisition easier, which further improves supply.

Starting this flywheel required investment in the unglamorous work of building robust, reliable tools for business users rather than the exciting work of building delightful consumer experiences. But once the flywheel started spinning, it became self-reinforcing.

The alternative—focusing on consumer features while stores struggled with clunky tools—would have meant stores gradually reduced their usage or churned entirely. All the brilliant consumer features in the world don't matter if there's nothing to buy when users open the app.

Learning From Store Behavior

Building tools for stores also generated valuable data and insights about how the marketplace actually functioned. Guillermo and the team could see which stores published consistently versus sporadically, which product categories sold fastest, what times of day had best conversion rates, and how pricing affected purchase behavior.

This operational intelligence informed everything else they built. When they eventually did return to consumer features, they knew exactly what mattered most because they understood the supply side deeply. They knew which neighborhoods needed more stores, which product categories had excess demand, and what times of day users most wanted specific types of food.

This is another compounding benefit of supply-side investment. Consumer feature development becomes dramatically more effective when you understand supply dynamics. You don't waste time building features for use cases that supply constraints make impossible. You focus on features that work with the reality of how stores actually operate.

The Sales Feedback Loop

Guillermo's background in B2B sales shaped how he approached store acquisition and relationships. He spent enormous amounts of time in conversations with potential and existing store partners, gathering feedback about what worked and what didn't.

"Obviously, they are not the best advisors, but they tell you the problem."

This hands-on approach meant the product team wasn't guessing about what stores needed. They had direct feedback from dozens of conversations each week. When a store said something wasn't working, the team could investigate the root cause and prioritize fixes that would have broad impact.

This tight feedback loop between sales, product, and engineering accelerated the pace of improvement. Instead of quarterly roadmap planning based on assumptions, they were shipping weekly improvements based on actual store pain points. The development team operated more like a services organization responding to client needs than a product team building for a theoretical user.

For early-stage companies, this responsiveness creates competitive advantages. Stores would try competitor platforms with slicker consumer apps but clunky store interfaces. Then they'd come to Buen Provecho and find that publishing products was actually easy, analytics were actually useful, and the team actually responded to feedback. That responsiveness converted stores and built loyalty even when the consumer app was less polished.

The Consumer Feature Backlog

During this period, consumer feature requests piled up. Users wanted better search and filtering. They wanted to save favorite products, not just favorite stores. They wanted more detailed product information and photos. They wanted delivery options instead of just pickup. They wanted social features to share finds with friends.

All reasonable requests. All features that would improve the consumer experience. All pushed to the backlog while the team built store tools.

"Now we're going back to the app and we are actually developing AR recommendations, personalized on what you have been buying, what you like, how we can basically get you to purchase more, frank, and always keep you as a user."

The transition point came once stores had robust, reliable tools. With 85-90% of development time spent on the supply side for nearly two years, stores could now efficiently manage their presence on the platform. Publishing products was straightforward. Analytics were clear. The system worked reliably.

Only then did the team feel comfortable shifting development focus back toward consumer features. They started building recommendation systems that could suggest products based on purchase history and preferences. They began work on personalized notifications beyond just location-based targeting. They explored ways to help users discover new stores and product categories.

But this consumer feature development benefited enormously from the foundation built through supply-side investment. The recommendation algorithms could work because there was enough data from engaged stores and consistent product flow. Personalized notifications mattered because there were enough products available to make them relevant. Discovery features worked because supply was substantial enough to make discovery valuable.

The Counter-Example: Competitors

Guillermo noted that competitors in the Latin American food waste space took different approaches, many focusing heavily on consumer experience with slick apps and sophisticated marketing.

"The competitors mainly like, I wouldn't say it in a bad way, but they kind of cloned the model, right?"

These competitors often had better-looking consumer apps. They had more polished onboarding flows. They invested in paid user acquisition to drive downloads. But many struggled with store retention and consistent supply because they hadn't invested proportionally in making the store experience excellent.

The result was predictable: impressive consumer download numbers but poor retention because users couldn't reliably find products. Stores would onboard, struggle with clunky tools, reduce their engagement, and eventually stop using the platform. The marketplace failed not from lack of consumer demand but from inadequate supply-side infrastructure.

Buen Provecho's willingness to prioritize the unglamorous work of building robust B2B tools over the exciting work of consumer features created a sustainable competitive advantage. Once stores loved using their platform, they published more products, which made consumer acquisition more efficient, which improved unit economics, which funded further growth.

Lessons for Two-Sided Marketplaces

The Buen Provecho story offers crucial lessons for any marketplace business trying to solve the cold start problem and build sustainable growth.

Identify which side constrains growth. In supply-constrained marketplaces, all the consumer growth in the world doesn't help if there's nothing to buy. Figure out which side is the bottleneck and invest there first. Build tools that remove friction from the constraining side since reducing friction even slightly can have dramatic effects on supply or demand depending on where the constraint lies. Create feedback loops with the constraining side through regular conversations, usage analytics, and rapid iteration based on what you learn. Be willing to disappoint the unconstrained side temporarily, knowing that solving the constraint ultimately serves everyone better than perfecting an experience built on shaky foundations.

For marketplaces using engagement and gamification features, the lesson is clear: engagement mechanics only work when the underlying marketplace has healthy supply. You can't gamify your way out of fundamental supply problems. Build the supply foundation first, then layer on engagement features that make a working marketplace even better.

The Path Forward

With nearly two years of supply-side investment behind them, Buen Provecho is now in position to accelerate consumer feature development while maintaining the strong store foundation they built.

The store dashboard is robust and reliable. Stores can easily manage their presence, track their impact, and see the value they're getting. This means the team can shift development resources toward consumer features without worrying that store tools will degrade.

The next phase involves sophisticated personalization, better discovery mechanics, and features that help active users increase their purchase frequency. These consumer improvements will drive revenue growth now that supply constraints are less acute. But none of it would matter if they'd built these features first while stores struggled with unusable tools.

Key Takeaways

• Focus development on the constraining side of the marketplace as supply constraints require supply-side investment regardless of consumer feature desires
• Store feedback loops accelerate product development through direct conversations revealing real problems beneath feature requests
• Supply-side investment creates compounding returns since more engaged stores publish more products, improving consumer experience and enabling organic growth
• Consumer features work better after solving supply because personalization and discovery need sufficient products to be valuable
• Willingness to delay gratification distinguishes sustainable marketplaces as the unglamorous work of B2B tools creates competitive advantages competitors won't replicate

Listen to the full conversation with Guillermo Martinez on the Levels Podcast to hear more about building the supply side of a marketplace and making hard resource allocation decisions.