A two-sided marketplace is only as good as both sides. The pharmacy side of what we are building at Covalenty gets most of the attention in how we talk about the product, because pharmacies are the buyers and the buyer experience is what most directly determines whether orders happen. But the distributor side is equally important, and probably more difficult to get right.
This article is about the distributor side: what we look for in a distributor partner, how we approach onboarding, what creates tension in the relationship, and what we have learned so far that has changed our thinking.
What Makes a Good Distributor Partner
Brazil has over 120 licensed pharmaceutical distributors. Not all of them are suitable partners for a marketplace platform, and the reasons have less to do with size than with operational fit.
The most important attribute is catalog quality. A distributor that can provide structured, accurate product data, including ANVISA registration numbers, lot numbers, expiry dates, and consistent product naming, is far more valuable than a large distributor that cannot export a clean product catalog. Manual data entry of distributor catalogs does not scale, and inaccurate product data creates real compliance risk for pharmacies. We have turned down relationships with larger distributors because their data management was too messy to work with safely.
The second attribute is fulfillment reliability. When an order is confirmed on the platform, it needs to actually ship. A distributor that accepts orders and then cancels a significant percentage of them for inventory reasons is actively harming pharmacy trust in the platform, even if the cancellation is not technically Covalenty's fault. From the pharmacy's perspective, a confirmed order that does not arrive is a confirmed order that did not arrive. The platform that facilitated it is the one they associate with the failure.
The third attribute is ANVISA authorization status. This is table stakes: every distributor on the platform must hold a valid AFE. We run this verification at onboarding and on a recurring cycle. It is not negotiable and there are no exceptions.
The Onboarding Process and Why It Takes Longer Than Expected
When we started signing up distributors, we underestimated the friction involved in the data integration step. Most distributors in Brazil's pharmaceutical market do not have a documented API or a standardized export format for their product catalog and inventory. Their data lives in internal ERP systems with varying degrees of structure, and extracting it in a format that maps cleanly to our catalog schema requires a custom integration for each distributor.
The custom integration work was expected to take a few days per distributor. In practice, it took longer for most of them. The issues were not technical in the sense of being architecturally complex: they were data quality issues. A distributor's internal naming for a product would not match the ANVISA nomenclature. Lot numbers would be formatted inconsistently. Fields that we needed populated would turn out to be optional in the distributor's internal system and therefore often empty.
We addressed this by building a validation layer that catches common data quality issues at the time of catalog ingestion, rather than waiting for a pharmacy to encounter a bad record when trying to place an order. The validation does not catch everything, but it significantly reduces the rate of bad data reaching buyers. More importantly, it provides structured feedback to distributors about the specific data quality issues in their catalog, which gives them something actionable to fix in their own systems.
Geographic Coverage and the Delivery Radius Problem
Brazil's pharmaceutical distribution geography is fragmented. National distributors can in principle reach any pharmacy in the country, but their delivery timelines and minimum order requirements for remote markets are often impractical for small independent pharmacies. Regional distributors can serve local markets more efficiently but have no presence outside their operating region.
Our initial distributor network was almost entirely concentrated in the greater Sao Paulo area, which reflects where we started and where most of our pharmacy relationships are. That is rational for an early-stage platform: go deep in one market before expanding. But it means that a pharmacy in Porto Alegre or Belem who finds us online is looking at a platform that currently cannot serve them well, and we have to be honest about that rather than overpromising.
We are not claiming to have solved the geographic coverage problem. We are working through it city by city, starting with São Paulo and expanding to other major metropolitan areas. A platform that tries to expand everywhere at once and delivers poor coverage everywhere is worse than one that delivers good coverage in a smaller geography first.
The Tension Between Distributor Interests and Pharmacy Interests
A marketplace that shows prices from multiple distributors simultaneously creates competitive pressure between them. A distributor who would prefer to be the pharmacy's only supplier has an inherent tension with being one of several options visible on a comparison screen. We have had direct conversations with distributor partners about this tension, and we do not try to pretend it does not exist.
The argument we make to distributors is that the alternative is not that they keep all their existing pharmacy relationships indefinitely. The alternative is that pharmacies continue their current pattern of consolidating around two or three preferred distributors and stopping active comparison. In that world, distributors who are not among the preferred set get no orders at all from a given pharmacy. In the platform world, a distributor with competitive pricing on a specific product category can win orders from pharmacies that would never have called them through a rep relationship.
This argument resonates more with smaller and regional distributors who are struggling to get their pricing in front of buyers than with large national distributors who already have broad rep coverage. The platform value proposition is not symmetric across all distributor sizes, which is something we knew going in and which shapes how we approach distributor recruitment.
What We Have Learned About Distributor Quality
The biggest lesson from the first months of distributor onboarding is that data quality predicts fulfillment reliability better than size or reputation does. A smaller regional distributor with clean, well-maintained product data and consistent lot tracking tends to have better order fulfillment rates than a larger distributor with a messy catalog. This is probably not causal in a simple sense: both data quality and fulfillment reliability are probably downstream of the same underlying operational discipline. But as a selection criterion, catalog data quality is something we can evaluate before signing up a distributor, whereas fulfillment reliability requires actual order history to assess.
The second lesson is that the relationship investment required to maintain a distributor partner is ongoing, not just front-loaded at onboarding. Catalog updates, new product additions, and changes in stock levels all need to flow into the platform in close to real time to be useful. A distributor whose catalog update process requires a manual export and manual upload once a week is providing information that is consistently slightly out of date. Managing the pace and quality of data refresh is a continuous operation, not a project.
