Every time I write an article about Low-Code, half of those commenting are writing comments that implies they are confusing low-code with no-code. My reasons for knowing this is because their arguments are as follows.
- Low-code is not flexible enough
- It is only suitable for MVPs
- You'll get stuck and you'll have to restart the project using "full code"
- Etc ...
Let me therefor define low-code by using the simplest possible analogy I can.
Low-code is just like "ng generate" or "ng new" from Angular - Except where ng stops, low-code frameworks keeps on adding thousands of additional lines of code to your solution
At least our low-code solution generates code. There is nothing mysterious about it. If you've used Entity Framework from Microsoft, you've arguably used low-code for years. When you click on your database in EF and choose "Generate model", you get thousands of lines of code, automatically generated for you, allowing you to start using your model from C#. When the code is generated, you can do whatever you wish with it. Edit and modify it, add to it, or delete parts of it for that matter. The difference is that when using low-code, the machine does all the boring parts - The parts that also happens to be extremely time consuming, repetitive in nature, and that you have to spend months to manually apply.
I demonstrated our low-code solution for several senior advisors at PwC in Cyprus a couple of days ago. I'll let the video speak for itself.
Pay particularly notice to how a senior software developer, in fact a CTO, with 15+ years of software development experience, told me he would need 8 to 9 months to reproduce what you can do with our low-code generator in 2 seconds. True story!
Low-code is only suitable for MVPs if you look at the result it produces as your finished solution. However, it's called low-code for a reason, and that reason is that you don't have to add a lot of code to it, but rather just a "little bit" of code. You can still add as much code to the result as you wish, allowing you to modify it any ways you want to. This implies that low-code will not get you stuck anymore than any other programming languages or frameworks will. Whatever you can do with "full code" you can also do with low-code. Whether or not your low-code framework of choice automatically generates that particular parts of your code or not, is a different story. However, let me repeat the above once more ...
You can do EVERYTHING with Low-Code that you can do with "full-code"!
If you believe low-code is only suitable for MVPs and simple apps, you've completely missed the point. Low-code simply gives you a head start, and that's it! Nothing mysterious about it. However, when that head start implies 8 to 9 months of manual labour, the economy of things starts changing the rules of the game.
HOWTO double your salary
In Cyprus, the average software developer is probably making some 5,000 EUROs per month. If my people can get a head start of 8 to 9 months delivering outsourcing services compared to our competitor, this implies we can deliver an application to our clients 40,000 EUROs less expensive than our competitors, and still have the same monthly profit.
If I showed you two identical cars, and I said you could buy one for 20,000 EUROs, while the other one would cost you 60,000 EUROs - Which would you choose?
As software developers we've been sheltered from the brutal facts of economy. Our managers have mostly been satisfied as long as we do whatever it is they want us to do, and paid our salaries happy that we've chosen to work for them. However, your manager is not ignorant to economy. If he can have some guy being able to deliver the same result in 2 months that some other guy needs 11 months to deliver, this literally translates into him getting an additional 50,000 EUROs in profit that year if he goes with the faster guy.
Who do you think your CEO would choose? The guy spending 2 months or the guy spending 11 months?
This paradoxically creates an opportunity for you to raise your own salary, simply because if you tell your manager you only need 2 months to deliver something others will need 11 months to deliver, he will be happy to pay you 10,000 EUROs in monthly salary, instead of 5,000 EUROs - Because even though you are more expensive on an hourly basis, he is still saving 30,000 EUROs on his final bill! Let me repeat that slightly emphasised ...
Low-code allows you to DOUBLE your salary, assuming you can quadruple your productivity, since the guy paying your salary is still saving 50% of his costs!
To illustrate the point of this article, watch the following video where I demonstrate what I was able to achieve by adding 5 days of manual coding on top of our low-code stuff.
What I did in 5 days of coding
First of all, we are administrating almost all aspects of our company through this system. Yup, you heard me correctly - This is the only internal administration system we are using in Aista. Below is a screenshot of the KPI dashboard.
In addition, the system provides integrations towards our hub, implying every time a new user is signing up at hub, it automatically becomes a contact in our CRM, allowing us to track communication, tasks, cloudlets, activities, etc.
In addition, we are using the system to send out newsletters to registered users, having confirmed their email while signing up, allowing us to directly market to them.
It is also a multi user CRM system, allowing us to administrate users and declaring what type of access each individual user has.
Of course, in comparison to SalesForce or Zoho, the system is simple - However, it provides us with hundreds of features neither SalesForce nor Zoho could ever provide us easily - Which are integrations with our existing custom systems, providing us real time data of how many users are registering, what cloudlets they are creating, giving support on support tickets, etc, etc, etc. From a functional point of view, we'd need at least 5 other systems to deliver what this one system gives us out of the box, and most importantly of course ...
One man created it in 5 days!
From a financial perspective, the system runs in circles around every single software developer on the planet still manually creating code. With our low-code platform, each of our individual developers becomes equally productive as 20 traditional software developers, making us 20 times faster, allowing us to invoice our clients 10 times as much, yet still be 50% less expensive.
Low-code is the future, everything else is "the abyss" ...
Top comments (4)
Thank you for an excellent write up regarding low-code, I concur that it is under appreciated and often misunderstood, but it has an insane amount of value towards deliverables, so long as you RTFM and play by the rules.
As I learned after several years of engineering, as much as I like writing code, I don't like writing all the details necessary to build a secure & robust application. I'd rather spend my time on true feature development when I write code.
Word!
Low-code is not suitable for everything. But where it fits, it allows its users to focus on what matters. I must have created a bajillion login forms in my life. Do I really want to create one more? Spending a week in the process? Low-code is just the natural evolution of library-based code. Once you've understood this much, you've understood everything you need to know 😉
Love the entity Framework analogy, I've always thought of LowCode as standardising and reusing code. The remaining bespoke code is where the value is and what customers are actually paying for
Word! In fact, it could be argued we've used low-code for as long as we've had libraries, DLL files, and reusable snippets of code. I guess the only difference, is that (some) low-code frameworks, allows to use meta information to "generate" new code, that somehow becomes library code, for you to use in your own projects.
However, there's nothing new under the sun really. Every time I explain meta programming I start out by telling people that AI was invented in 1948 (Perceptron), but it was "dead" for 50 years. This is called the AI "winter". During this time, there was tons of research conducted on the ideas though. Meta programming (my field) has a similar foundation back in the 1960s and 1970 with 4GL and Lisp, but was then "dead" for 50 years, until I started reanimating it a decade ago with Hyperlambda.
The point being that everything we think is new, is really very very old. MVC for instance was "reinvented" with RoR in 2005. However, it was invented in 1979 actually ...
Everything is like that. All the good parts of software development, are really old constructs, being reinvented and rediscovered by some smart dude, figuring out it's useful for some reason ...