Why 38% of Purchases Come from Favorited Stores (And The Problem It Creates)

When power users love your product so much they favorite every store and snap up inventory instantly, that sounds like a retention win. But for Buen Provecho, a food waste marketplace operating across Latin America, this feature created an unexpected problem: their most engaged users were making it nearly impossible for new users to find value.

On the Levels Podcast, Guillermo Martinez, COO and co-founder of Buen Provecho, explained how a seemingly successful engagement feature exposed the tensions inherent in supply-constrained marketplaces—and why optimizing for your best users can sometimes hurt overall growth.

The Favorites Feature

The concept seemed straightforward: let users favorite stores they love, then notify them immediately when those stores publish new surplus food. It's a pattern borrowed from countless apps—favorite content creators, save searches, follow brands. Users get what they want, engagement goes up, everyone wins.

For Buen Provecho, the feature worked exactly as intended. Users who favorite stores demonstrate clear intent—they've found places they trust, products they enjoy, and locations convenient to them. When notified about new inventory, they act quickly.

"We have users that really like some particular stores and you can like favorite them. So you get a notification in the second that they publish something. So that actually is like a double-edged sword, right? Because active users get notified immediately. So they purchase instantly."

The engagement metrics looked incredible. A staggering 38% of all purchases came from users' favorited stores. That's not a typo—more than one-third of their entire transaction volume happened because users had explicitly chosen to follow specific stores and bought the moment they got notified.

The Second-Tier User Problem

But here's where the success metric revealed its downside. New users—those without favorites, those still exploring, those who haven't yet found their go-to stores—were essentially competing for inventory against users who got instant notifications.

"But if you're new, you get kind of screwed."

The dynamic creates a two-tiered marketplace experience. Imagine you're a new user browsing the app. You see an interesting item, click through to learn more, decide you want it, add it to your cart—and by the time you hit purchase, it's sold out. Someone who favorited that store got notified five minutes earlier and bought it immediately.

Do this a few times and new users conclude the app doesn't work for them. They're always arriving too late. Meanwhile, power users keep purchasing successfully, reinforcing their behavior and engagement.

Why This Matters for Retention

The favorites feature inadvertently created a retention barrier for new users during their critical first experiences. Early adoption research consistently shows that users need to experience value quickly—often within the first session or two—or they churn.

When power users snap up inventory instantly, new users can't have that successful first purchase. They open the app, see interesting products, try to buy, fail to complete the transaction, and associate the app with disappointment rather than discovery.

Guillermo recognized this creates a longer-term problem for growth. Yes, the 38% of purchases from favorites demonstrates strong engagement from existing users. But if new user activation suffers because of this dynamic, the overall user base stops growing even as existing users remain active.

It's reminiscent of the classic engagement versus acquisition trade-off, but with a supply constraint twist. Most apps can serve both engaged users and new users simultaneously. Buen Provecho's limited inventory forces them to choose who gets access first.

The Marketplace Dynamics

This challenge reflects a broader truth about marketplaces: features that work well in abundant supply environments can break down when supply is constrained.

Consider e-commerce. Amazon can let users save searches and get instant notifications about new products because their supply is effectively unlimited. If ten thousand users want the same book, Amazon ships ten thousand books. There's no scarcity.

But in a marketplace with limited, variable inventory—whether it's surplus food, event tickets, rental properties, or freelance services—instant notifications to favorites create winner-take-all dynamics. The fastest users with the best notifications get everything. Everyone else gets nothing.

The favorites feature essentially weaponizes speed. Power users who've invested time learning the system, favoriting all their preferred stores, and keeping notifications enabled have a massive structural advantage over newcomers still figuring out how the app works.

Potential Solutions

Guillermo and his team are actively thinking about how to address this imbalance. Several approaches could help, though each comes with trade-offs.

Reserved inventory for new users. Allocate a percentage of each store's surplus specifically for users who don't have that store favorited. This ensures new users can actually complete purchases during their critical first sessions. The downside is disappointing power users who expect instant access.

Graduated notification timing. Instead of sending instant notifications to all users who favorited a store, stagger them slightly. Send to a subset first, then expand after a few minutes. This slows down the race to purchase and gives more users a chance to discover items. But it also delays gratification for engaged users and could feel arbitrary.

Purchase limits from favorites. Cap how many items from favorited stores a single user can buy in a given time period. This prevents power users from monopolizing inventory. However, it punishes your best customers and could reduce overall sales.

Better matching algorithms. Rather than first-come-first-served, use algorithms to match products to users based on purchase history and preferences. This could distribute inventory more fairly while still rewarding engagement. The technical complexity is significant, though.

Gamification with fairness mechanics. Implement leaderboards or achievement systems that reward both power users and new users differently, creating separate incentive structures that don't directly compete for the same inventory.

Each solution involves choosing which users to prioritize: the engaged power users who drive most transactions, or the new users who represent future growth. There's no obviously right answer.

The Broader Lesson

Buen Provecho's favorites feature demonstrates an important principle for product builders: the same feature can be simultaneously a retention driver and an acquisition barrier.

The feature works beautifully for its intended audience—users who've already found value and want more of it. The 38% transaction rate proves that. But its very success creates challenges for users who haven't yet reached that stage of engagement.

This is particularly acute in supply-constrained environments. When there's not enough to go around, features that help some users get more access inherently mean other users get less access. Zero-sum dynamics force difficult prioritization decisions.

Many product teams would look at "38% of purchases from favorites" and declare victory without digging deeper. The metric seems unambiguously positive. Only by talking to new users—those struggling to complete their first purchase—does the downside become visible.

What This Means for Growth Strategy

The favorites feature highlights why Buen Provecho's growth strategy centers on adding more stores rather than more users. With more inventory, the tension between power users and new users decreases. There's enough surplus food for instant notifications to favorites while still leaving discovery opportunities for newcomers.

Guillermo mentioned they're focusing heavily on onboarding large chains and franchises specifically because this unlocks supply at scale. Sign one partner with 30 locations and suddenly the favorites problem becomes less acute—there's simply more inventory to go around.

"We needed to go to big stock stores, retailers, supermarkets, those kinds of things."

This supply-first approach means temporarily accepting lower new user activation rates while building the infrastructure to support better activation later. It's a patient strategy that prioritizes sustainable growth over quick wins.

Key Takeaways

• Success metrics can hide problems when 38% of transactions come from one feature, dig into who's NOT using that feature successfully
• Power user features create barriers in supply-constrained environments by giving advantages to experienced users that newcomers can't access
• Instant notifications amplify scarcity by turning inventory availability into a speed competition that favors users who've optimized their setup
• Trade-offs are inevitable between optimizing for engaged users versus new users when supply is limited
• Supply solves feature problems more inventory reduces zero-sum dynamics and allows features to work for everyone

Listen to the full conversation with Guillermo Martinez on the Levels Podcast to hear more about how Buen Provecho balances power user engagement with new user acquisition in a supply-constrained marketplace.