How Model Car Collector Grows Organically with Product Optimization and User Feedback

My cousin has a collection of over 5000 diecast models, and in 2022 we had a discussion around apps he tried to use for storing and managing his collection, but all fell short in some way or another and didn't cater to his needs. Most of the tools we found were either too basic, focused on selling, or not built around the details collectors actually care about. Diecast collectors want to track things like manufacturer, series, scale, model year, condition, photos, storage location, and value, and I felt there was room for something purpose-built. I also had a few diecast cars and wanted to learn a mobile platform, so I decided to build an app that could help us out and at the same time learn a new technology. I started Model Car Collector as a relatively simple Android app (released April 2023) and kept adding features as I learned what collectors were actually using and asking for. Over time it grew into something much bigger than the original idea, with collection analytics, public profiles, albums, social features, valuations, and eventually AI-assisted model identification. The turning point for me was realizing this was not just a small utility I had built for myself and family. People were downloading it, building large collections in it, and relying on it to keep track of something they had spent years collecting.

The growth tactic that has worked best has been improving the product and its Google Play presence consistently rather than relying on a single big marketing push. Model Car Collector has grown to tens of thousands of Android downloads, with much of that coming organically through Play Store search and discovery. Updating screenshots, descriptions, categories, onboarding, and responding to reviews has made a noticeable difference over time. What hasn't worked well is assuming that simply adding more features automatically creates growth. As a developer, it is very easy to spend weeks building something because it feels useful, then discover that users barely notice it. I have learned that distribution, positioning, and explaining existing features can matter just as much as building new ones. Some of the most useful changes have actually been relatively small improvements to the store listing or user experience, while some much larger development efforts had far less impact on downloads or subscriptions than I expected.

The biggest complaints have usually been around limits, pricing, and expectations about what should be available for free. Earlier versions of Model Car Collector limited how many models a free user could add. From a subscription point of view, that seemed reasonable, but collectors understandably saw cataloging their collection as the core purpose of the app. Some users had hundreds or even thousands of models, so hitting a hard limit created frustration and negative reviews. I eventually removed that limit and changed how I think about monetization. Instead of restricting the basic reason someone downloaded the app, I now try to make the paid tiers valuable through additional features such as AI identification, valuations, analytics, and other advanced tools. I also read and respond to user feedback regularly, even when it is uncomfortable. Not every request can be implemented, but repeated complaints are usually a signal that something in the product or the way it is being communicated needs attention.

I think most people underestimate how serious diecast collecting can be. From the outside, it can look like people are simply buying small toy cars, but many collectors have hundreds or thousands of models accumulated over the years. They care about specific manufacturers, castings, series, variations, scales, model years, condition, packaging, rarity, and value. A serious collection can contain a surprising amount of information. That is also why generic inventory software is not a great fit. Collectors do not just want to know that they own three cars. They want to know exactly which versions they own, where they are stored, what they paid, what they may be worth, and what is still missing from a set. Another misconception is that valuations can always be reduced to one perfect number. Collectible markets move constantly, and condition, region, scarcity, and recent sales all matter. The real challenge is giving collectors useful information while being honest about uncertainty.

The biggest next step is bringing Model Car Collector to iOS so the platform is no longer Android and web only. I also want to keep improving the AI identification and valuation features, especially the quality and confidence of the information returned to collectors. On the web side, the focus is on better analytics, market insights, collection sharing, and community features. The goal over the next year is to make the product feel like one connected platform across mobile and web while continuing to grow the collector community internationally.

65,000+ lifetime Android downloads, over 125,000 diecast models uploaded by the community, and containing over 500,000 images in total.

Before building Model Car Collector, I was already an experienced software engineer, mainly working with .NET, databases, web applications, and cloud systems. I had spent years building business software professionally, so I was comfortable with the technical side of creating and maintaining a real product. What I did not have was experience running a consumer subscription app or building specifically for the diecast industry. A lot of the product, marketing, and monetization sides have therefore been learned while building the company. I also moved into Flutter and mobile development as the app evolved, and later built the web SaaS and API around it. Model Car Collector has been a very different challenge from normal client or enterprise software because users decide very quickly whether a product is useful enough to keep installed, recommend, and eventually pay for.

One of my biggest mistakes was putting too much emphasis on restricting the free product as a way to drive subscriptions. At one stage, free users could only add a limited number of models. On paper, it seemed like a logical conversion point, but in practice, I was limiting the main reason collectors had downloaded the app in the first place. It generated frustration and some negative reviews. I eventually removed the collection limit and shifted the paid value toward features that genuinely add more capability, such as AI identification, valuations, and analytics. The lesson was that monetization works better when users want to upgrade because the premium features are valuable, rather than because the core product has become inconvenient. I have also learned not to become defensive about criticism. Even an angry review can contain useful information about where the product or pricing model is creating unnecessary friction.