Building Became Easy. Getting Noticed Became Almost Impossible

For a few years, a perfectly reasonable fantasy existed in the software industry: all it took was a good idea, working hard to build a website/app, and putting it online or in the app store. If it was truly good, people would eventually find it. A single creation could change the lives of its founders.

Cases like Instagram, YouTube, etc. turned this possibility into a founding myth. A small team could launch a product, win over millions of people, and create a huge company in very little time. The story was extraordinary, but it seemed reproducible enough to guide an entire generation of developers.

The video “AI Killed The App Economy”, by Kenny Gunderman, begins with this promise and reaches a much less comfortable conclusion: it has never been so easy to build an app, and perhaps it has never been so difficult to turn one into a business.

The problem is no longer the code. It is distribution and trust.

When the store handled distribution too

In the early days of app stores, there was a combination that is difficult to reproduce. Millions of people had just bought a new kind of computer, carried that computer in their pockets, and wanted to discover what it could do. At the same time, there were relatively few apps available.

Software was scarce and difficult to develop. Attention and demand, meanwhile, were enormous.

The App Store itself functioned as a distribution channel. Appearing on a list, reaching the top positions in a category, or being mentioned on a technology website could generate a huge wave of downloads. Users had not yet chosen an app for every need. There were empty spaces on the screen and in the imagination.

It was a real gold rush, but a short one. As categories gained leaders, those spaces disappeared. The winners accumulated users, reviews, data, advertising budgets, and brand recognition. A new app no longer needed only to be good. It needed to convince someone to abandon a familiar solution, learn another interface, and trust their data, or their money, to a company they had never heard of.

AI removed the last barrier

There was still one protection left: building software was expensive and laborious. You had to learn how to program, hire a team, or raise money. Even a simple idea required months of work, infrastructure, and maintenance.

AI is undoing that protection. In the video, Gunderman says that he modernized a personal app with around 20,000 lines of code in one night while playing RuneScape. The case does not mean that anyone can create any system without experience. It means something more important for the market: the minimum cost of producing acceptable software has fallen dramatically.

When millions of people can turn an idea into a product, having a product ceases to be an advantage. Every new app enters an already crowded store, surrounded by competitors that have also become faster and cheaper to build.

The data shows the scale of the change. RevenueCat’s State of Subscription Apps 2026 report analyzed more than 115,000 apps, $16 billion in revenue, and more than one billion transactions. According to the report, approximately 2,000 new subscription apps were launched each month three years ago. Now it is almost 15,000.

Demand did not increase sevenfold. Supply did.

The market is growing, but not for those arriving now

At first glance, there is still a great deal of money circulating. Sensor Tower estimates that in-app purchases generated $167 billion in 2025, an increase of 10.6% in one year. Downloads, however, grew by only 0.8%.

This describes a mature market: more revenue extracted from a base that is barely growing. And that revenue is not being distributed evenly.

In RevenueCat’s study, apps launched before 2020 accounted for 69% of all subscription revenue in January 2026. Those launched in 2025 or later, already in the era of vibe coding, accounted for only 3%. The older group had years to accumulate an audience, improve monetization, and turn growth into more growth.

Even among apps that made it far enough to enter the report’s dataset, only 17.3% of new ones reached $1,000 in monthly revenue in their first two years. Only 4.6% reached $10,000 per month. The report also found a growing divide: the top quartile increased recurring revenue by more than 80% in one year, while the bottom quartile shrank by 33%.

These figures do not literally prove that the chance of creating the next Instagram is one in a million. RevenueCat’s dataset also does not represent every B2C product: it focuses on subscription apps and excludes projects that are too small to produce statistically useful data. Even so, the pattern is difficult to ignore. The easier it becomes to launch, the more the outcome concentrates among those who have already acquired what cannot be generated by a prompt.

Distribution is not an afterthought

Developers like to think of marketing as the next step: first we make a good product; then we find the users. That separation may never have been very real. Now it has become fatal.

Distribution is the ability to reach the right people repeatedly. It can come from an audience of your own, a community, an email list, a sales team, partnerships, physical stores, advertising, or another product that is already known. It is not just getting a spike in views. It is being able to show up again tomorrow without starting over from zero.

Trust is what converts that presence into a choice. It comes from familiarity, recommendations, reputation, reviews, support, and the perception that the product will continue to exist. For a consumer, switching apps does not cost only the subscription price. It costs time, data, habits, and the risk of betting on something that may disappear.

Established companies start with both resources. They can show a new product to current customers, buy attention at scale, use a familiar brand, and sustain months of experiments. A newcomer starts without customers, reviews, or recognition.

That is a difficult opponent to beat with more features alone.

Starting a B2C business from zero has become a lottery

A product for businesses (B2B) can survive with few customers. It is possible to identify buyers, talk directly to them, understand an expensive problem, and close contracts one by one. Selling is hard work, but the path to the customer is visible.

In B2C, the math is different. Low prices require thousands or millions of people. There is no practical list of consumers to call. Each individual needs to discover the product, understand it, trust it, and change their behavior. Doing this at scale means competing with companies that already occupy the public’s screens, memories, and routines.

That is why launching a B2C product from zero in the hope that quality will produce discovery is, in practice, buying a lottery ticket. Every now and then, an extraordinary case appears, precisely because extraordinary cases are the ones that receive coverage. We do not see with the same frequency the thousands of competent apps that never found their first hundred users.

The next Instagram may exist. But “being the next Instagram” is not a reproducible strategy. It is the name we give the winner after an extremely rare combination of product, timing, social behavior, network effects, and distribution works out.

For those arriving now without an audience, partners, capital, or a channel of their own, broad B2C competition is close to impossible. A newcomer does not compete only against better or worse products. It competes against years of accumulated exposure. And AI, by multiplying the number of alternatives, makes its position worse: each new option increases the noise and drives consumers even more toward what they already know and trust.

The only possible entry begins before the product

This does not mean that new consumer businesses are over. It means that the old path, build first and distribute later, is practically over.

A new product can still emerge from a community that already trusts its creator, from an audience cultivated over years, from a very specific need ignored by larger platforms, or from a partnership that provides immediate access to the right users. It can start so narrowly that it does not seem threatening and grow with the reputation earned within that niche.

But notice what all these paths have in common: distribution already exists or begins to be built before the software. The product stops being the origin of the business and becomes the way to serve an existing relationship.

The business is being chosen

For a long time, programmers could treat sales and marketing as the unpleasant work that would begin after the app was finished. Now they are the business itself. Software is necessary but abundant. Attention backed by trust is scarce.

This reversal changes the first question of a project. Instead of “can I build this?”, we should ask: “why would someone choose this, and how will that person know it exists?” If the only answer is the app store, a cheap ad, or the hope of going viral, there will probably not be a business, but only a product.

Code has become easier. The human part has become more valuable. In the new B2C software economy, winning is not managing to publish. It is earning the right to be noticed, remembered, recommended, and chosen. In the end, as so often happens, it’s all about relationships.