Don’t build platforms and products at the same time

Building a platform alongside a product is hard. It is even harder when it is done for a brand new product. It requires a deep level of understanding of what a customer is looking for. Creators must imagine a system that will enable that with flexibility for pivots. Early iterations set the pace for creating the moat that will help the product be competitive.

During a debate, a friend asked me to explain further. “Isn’t it fundamentally going to help everyone if we build a platform for our new product?”

We worked through a metaphor to understand this better.

Imagine you are a new fruit salad company. You hire 6 people, ask 5 of them to make 5 great tasting, instagram worthy, super tasty salads that people would rave about. You mark 1 person to supply the fruit others will assemble and mix. This one person is immediately a bottle neck because they have a lot to deliver. So they go for the lowest common denominator option, cut everything into cubes. They will only use fruits that are easy to peel — no pineapples. The juicer they got doesn’t work well for leaves so those are out for the time being too.

The constraints imposed very early in the process have many downsides. It closes off possible outcomes due to scale problems. Two major issues:

1. We create a bottleneck in our supply chain so early in the process by adding a platform layer. Because we are thinking of the problem at scale, when scale doesn’t exist, we are creating a lot of redundancy. In turn, these redundancies slow everyone’s progress. Perhaps everyone’s waiting on apples while mangoes are being cut that only one dish needs.
2. Quality suffers as there wasn’t much of an opportunity to find something customers would love. At the start of a journey, all attention needs to be given to what feels right, what solves the actual problem and what are people willing to pay for. As we divide our attention towards other tasks, products start to get diluted and performance becomes lackluster. 5 products with the same 5 fruits would not create enough appeal.

It is much easier to scale something you deeply understand. Having something customers love already creates a good opportunity to build something strong to scale. In our fruit salad example, this means letting each person create their unique salads first—finding what brings lines of customers daily. Only when the demand becomes stressful should you step back and identify genuine patterns. Maybe you discover everyone’s using apple juice, and that specific component can be centralized. Or perhaps there’s no overlap at all, and the solution is simply hiring another specialist. The key is that these decisions come from real experience, not speculation. Scaling is, after all, part of the optimization phase of building something—it works best when you’re optimizing something that already works.

To be clear, I’m not arguing against all shared infrastructure. Buying fruits from the same seller reduces costs without limiting creativity. Tracking purchase dates ensures quality across all products. These are operational efficiencies that don’t constrain innovation. The problem arises when we abstract the core value proposition—when we decide all salads must use pre-cut cubes before we know if customers even want fruit salad. Share the boring stuff early if you must, but keep the creative core independent until you truly understand what works.

Hardest part is making something people love. Then try to scale it.