Bottlenecks in Codeless Infrastructure
What No-Code Needs to Work on Over the Next Decade
The no-code revolution has definitely taken the startup ecosystem by storm.
With more “no-code tool for _____” being announced every week, cohort-based programs to teach no-code building, and people creating cool products out of these tools, we’re moving towards a time where starting any kind of product or service just takes a couple of clicks.
These tools play a key role in democratizing creation, allowing not just coders to build cool, impactful products.
“It’s never been easier to build software. Gone are the days of needing to be a full-stack developer or engineer to build a startup. The proliferation of open source, low-code, no-code tools across categories are enabling a new generation of entrepreneurship across backgrounds and experience and closing the skills gap and making it easier to build programs with little programming experience. These tools and open source projects that are community developed and free for all to use will become the foundation for which commercial products are built.” — Villi Iltchev and Frances Schwiep
However, as codeless software becomes more dependent solely on no-code infrastructure as they scale, a couple of bottlenecks begin to unveil themselves.
Can’t solve every problem
“Software developers translate between what humans want and what computers can do. The second part — what computers can do — is a well understood domain that can be modeled beautifully, and tasks involving it can be automated trivially. The first part, what humans want, is a profoundly complex anti-inductive domain that every human in history has spent their whole lives trying to understand, failing to but scratch the surface.” — adrusi on HN
Although no-code infrastructure is great for a small set of foundation needs, solving more complex user problems as your product evolves will no longer fit in codeless “cookie cutters”.
A reaction to this has led to more low-code tools being created and no-code infrastructure evolving into low-code.
Poor pricing by volume
Following my previous thought, I believe middleware software, like Zapier, will dominate the no-code/low-code spec for the next decade. However, the cost to use these products sucks.
Products that easily connect product A to B with over 3,000 other apps to choose from will be what stays from the no-code/low-code boom. Using Zapier even helps software developers build MVPs rapidly.
However, what I’ve observed, using tools like Zapier and many others, is that the pricing (based on volume) is highly disadvantaged to that of costs for infrastructure built from scratch as your product scales.
One risk that I consider when looking at startups is “what other technologies does their product depend on in order to be successful?”
Softwares like AWS, Google Cloud, and the Twitter API have had a proven track record of 99.9% uptime, which is necessary considering the types of clients that depend on them.
However, there has yet to be clear confidence in products that depend on no-code/low-code infrastructure because of the number of external dependencies they depend on to keep them running.
What I’ve learned more and more is that being able to code is an incredible superpower today and it’s a skill that I hope will continue to be learned instead of just looking towards these no-code softwares.
At the end of the day, I believe that the best, most sustainable products today are built with code and it will continue to be that way.
As for no-code, I expect it to continue to have a place in providing people the ability to build “things that don’t scale” to prove their hypotheses about things they're curious about.