10
39 Comments

The No-code Fallacy

Reading so many posts on no-code tools. Yes, if you can't code they are good. Even if you can code, you can setup a simple landing with email form (eg PH Ship or Mailchimp Website) faster than when coding yourself. No-code is even better when you want non-tech people to maintain content on your site and have an easy access to your site (eg webflow is great here).

But

  • If you need a little extra you are so much faster when coding yourself
  • Learning to code takes time but learning to setup a simple landing with an email form, eg. in react and nextjs doesn't take months AND MOST IMPORTANT you set the foundation to get a basic understanding and from there it's six to 12 months to be able to build anything and eg running full-fledged kubernetes clusters yourself
  • People think they need to learn html, css, js, jsx, bash, vim whatever to code; you don't need to learn anything; just follow the most basic example of again eg nextjs to setup a simple landing page; you will learn the essentials of many fields and only learn more in each field when needed; no need to read books about JS; if you don't get something just accept and move on, at some point you will
  • Just think about what do you want to achieve. If you want to build a company remotely related to software you need to know how to code, otherwise your dependency to an expensive resource (coders) is too high and worse, you stay dumb
  • Finally, can you build a scalable business with little investment without software involved?

Don't get e wrong. In a perfect world it's better that everyone focus on their core skills and if yours is not coding then ok but in the early stage of a company someone has to code. Most of the work in that stage is coding and getting good coders (who are better than a no-code tool) in that stage is almost impossible.

posted to Icon for group No-Code
No-Code
on January 26, 2020
  1. 6

    I have nothing against No-Code, but history tends to repeat itself. For those of us that learned how to hand-code HTML back in the 90's, we remember the plethora of WYSIWYGs out there. Hell, even Adobe capitalized on it with DreamWeaver. I also really like how No-Code lends itself to the DIY ethos and there is absolutely nothing wrong with that at all. Here's the thing: good designers are good designers; good developers are good developers, rarely shall the two meet in a single individual. For example, how many professional musicians are also professional-level songwriters? Few and far between. So, what the No-Code movement will do is provide coders with a cottage industry of building all of the tools for the no-coders to use. If something scales way up, there is a great likelihood that v.2.0 will be built out with code. Developers don't have to worry about not having work; they just might have to shift from what they are currently building to building this subset of tools for no-coders. This is no different than the 90's except there are (a) way more people online thanks to smart mobile devices and (b) greater technology to make things look a lot better than they did (remember frames, anyone?). It will look better than what a WYSIWYG could pull off from 20 years ago, but the functionality will still be limited.

    1. 2

      i disagree. what you describe is a mainstream notion but great eg UIs are built by an intersection of coder and designer. You can do so much in terms of design as a coder than a dumb designer with his/her gui tools (adobe and friends)

      Frontends are too technical. A designer can make you a CI and CD and an easy landing page but something like tiktok? Never. just for the record, tiktok uses so much advanced tech (heavy prebuffering of multiple streams ahead, 60fps) in it frontend which are not about design but coding but again lead to good ui.

      the world is not black and white.

      1. 1

        @Cass1 I'm confused as to what part of my post you disagree with?

        1. 1

          Here's the thing: good designers are good designers; good developers are good developers, rarely shall the two meet in a single individual.

          1. 1

            That’s 100% true. Sure, there are exceptions, which is why there are IDEs, but it isn’t the norm. It’s not a knock on either, but a mutual respect for both. You are a little flippant toward designers and you really shouldn’t be. Anyone that thinks design is somehow ‘less than’ has never properly designed anything. Some folks are born with a golden ratio in their eye, some are not. Design principles can be taught, but good design is a matter of gut feeling.

            1. 1

              That’s 100% true.

              No it's not. Again, a designer couldn't even articulate or spec a product like TikTok. TikTok's UI is not about colors, shapes but about tech nobody else did before (see my other post), this is hard-core tech which comes from coders and NOT designers which again leads to good UI.

              You view of the world is limited and yes, it's good enough to build some boring landing page or another SaaS but not enough to innovate.

              1. 1

                Alright, so we've established that you are a contrarian, aren't a designer and aren't a coder (your own admission, not mine). I'm trying to figure out what exactly you are the authority on since your own opinion appears to obfuscate your ability to understand any perspective other than your own.

                So, what's the deal?

                P.S. The short-form video app you seem to be in love with has over 300 openings for positions outside of development. Hmmm, I guess it takes more than just a developer to make an app work, right?

                1. 1

                  I'm trying to figure out what exactly you are the authority

                  there is no need to answer this q because I reasoned my POV multiple times with--I think--good examples.

                  The web is more than 2 decades old, mobile 1 decade. You need more to innovate and make money. if your mindset is locked into the last 20 years (designer are designers and coders are coders), good for you but not my world.

                  re your comment on tiktok: this company is now a later stage company, so i dont know what's the message of your comment

                  cmon lets stop here and agree to disagree. I am sure there will be opportunities to be more aligned on IH

                  1. 1

                    Agreed, and I'm looking forward to it.

  2. 4

    There's a place for no-code tools. I've been running a small app on Bubble since 2015. I make about $3k a month, which is fine for a side project. I have no aspirations for this specific project being a full-time job.

    I built it entirely without code, and the customers are happy with it. That's all I need, so I don't see any advantage that would've come from building it with code.

    If I were trying to build a super scaled product to handle millions of users, I'd still start with a no-code platform like Bubble or Adalo. The app is going to get rearchitected frequently as it grows anyway, so might as well get users and functionally validated as quickly as possible for as little money as possible.

    I work for a massive software firm (VMware) and agree that there will always be a place for well-written code and engineers, but I don't think it's fair to generalize that for every creator and every purpose.

    1. 1

      Exactly... I just don't get all the hate on nocode...

      1. 2

        Well, I kinda do. Folks are positioning nocode vs traditional development and we're at the stage of tooling where people are building lots of junk really fast, but it's "neat" because it was "nocode."

        Nocode, to me, is just a tech stack. You use a tech stack to build whatever product needs to exist. For me, I choose a nocode tech stack because I'd rather focus on building product than learning how to code. It's a good choice for me, but I'm not going to pretend it's anything more than that.

        1. 2

          of course, but there are also lots of junk being built with code as well :) But totally agree, would say both serve different purposes. I just know I can rapidly experiment with ideas and if somethings picks up well I can move into a proper infra.

          1. 1

            Yep, agreed. For me, the difference is how the nocoders are communicating about their junk lol. There are a handful of loud voices saying silly things.

    2. 1

      The app is going to get rearchitected frequently as it grows anyway

      This is a deep point that often gets missed by us developers.

      In the old days software was so expensive to build and fragile that you had to architect it carefully from the beginning so it didn't explode. Web applications are not like this any more, you can spin one up in no time and refactoring and rebuilding has become much cheaper.

      The other thing is that devs assume they will be the only developer into the future. If your business grows you can always hire developers to re-build some or all of your stack. Several large pieces of successful software (Stripe is one that comes to mind) were re-built more than once.

      1. 2

        A big part of my day job is helping large firms modernize the existing applications. They built massive monoliths with multi-year project plans in the early 2000s and are now having a heck of a time scaling and updating their apps. We're helping them implement architectures that are designed to change and evolve so they avoid the sins of the past. It's wild stuff and just as expensive, if not more, than rebuilding a no-code product would be.

    3. 3

      This comment was deleted 3 years ago.

      1. 1

        It’s an edtech app for early childhood literacy called Joyful Literacy. Sells exclusively to schools and school districts, so there’s no self serve signup. I do the on boarding and stuff myself on their behalf.

        1. 1

          This comment was deleted 3 years ago.

          1. 1

            Its twice my average computer engineer salary in Brazil - to give you some perspective.

          2. 1

            No complaints here. I share it with the woman who created the content and to cover the costs of my other side projects, so it's not really going into my pocket.

            There are folks running serious businesses on Bubble, though. One of these days, maybe I'll be one of them!

  3. 2

    You have no products listed, literally nothing to show for your programming background. I bet if you were trying some no-code solutions you’d have something up and running by now.

    1. 1

      What this guy said.
      Personally I code everything myself except for stripe and twilio, but you gotta launch first.

  4. 2

    "eg. in react and nextjs"
    "running full-fledged kubernetes clusters"

    Yeah, your examples are not easy at all... I assume this is your tech stack, but the easiest way to go about programming for a non-tech person will be Laravel/PHP and Jquery, eventually Django/Python and Jquery.
    You added an extra level of difficulty to something, that explains how easy it is to start.

  5. 2

    Recently, I set up a no-code Wix site for basic validation of an idea, which worked very well.

    Beyond that, to have a functional MVP would have required functionality not available in the Wix platform, or would have been more time-consuming, expensive, and accruing of technical debt, than simply coding a custom solution (I am a developer).

    So, even for a developer, there is a place for no-code on the validation-to-MVP spectrum, depending on the nature of the proposed solution, and one's own skillset. Even if one is not a coder, at some point no-code will become more expensive than hiring a coder. That tradeoff of functionality and technical debt is hard to judge for non-coders, so I recommend regularly consulting with a coder adviser to get a feel for when that point is reached.

  6. 2

    +1 to what Cass1 wrote. Very well said.

  7. 1

    I've been thinking a lot about this problem and decided to add coding features to my no-code project (NoCo.io). I love that I can drag and drop most of the app, but the second I couldn't, I just abandoned the platform (my own platform!). That's part of the reason I decided to include the ability to write custom react components, and pretty soon, adding custom APIs as well.

  8. 1

    I am currently building out a Job Board Template, and I can tell you no-code is damn good. Sometimes I don't understand why people think no-code is bad, it's actually just a layer added above the code, like Windows when you start up your computer.

    The added layer visualizes the code, at the end of the day it's code!

    1. 2

      but you are 100x slower with guis

  9. 1

    I think you missed the point. I have written this before and it's not the first time this appears on IH. NoCode movement is not here to replace developers or take their jobs or anyhow imply that code is not necessary. It is here to make rapid prototypes faster. In the early stage of product development, it is very likely that you are probably building something that people don't want and you need to iterate. You can literally achieve an MVP level prototype in 24 hours. That's why also a lot of designers are using Webflow because it's just that much faster and convenient. The back-end part is just connecting other tools with a relational DB.

    You are right about this:

    • NoCode has some severe limitations (I am looking for workarounds on a daily basis, yet that also makes you resourceful and finding ways how to solve for such problems similarly to googling stack overflow => it's pretty much the same experience). Not every feature will be available to you, but that's also the fallacy that most developers go through that the value is in features. Yet, we know that the invisible hand of the market usually proves them wrong fairly quickly.

    And IMO wrong about points such as:

    • "Just think about what do you want to achieve. If you want to build a company remotely related to software you need to know how to code, otherwise your dependency to an expensive resource (coders) is too high and worse, you stay dumb"

    I am sorry but this is BS.

    1. The point of code is being the right tool for the right purpose - scaling, automating, improving. "Do things that don't scale " first, remember?
    2. Stay "dumb"? Huh? If you think that all these tools have no learning curve then you are largely mistaken. They still require some knowledge of design (grids, paddings, containers, divs etc.), dataflows (merging tables, transforming data, JSON formatting etc.), understanding relational databases (CMS) etc.

    Of course, if you want to scale you will need to write custom code but at this point, you are better off getting a friend as a co-founder or just hiring an employee without giving out equity. You can scale tools to $10k MRR and many projects have proven that in some cases even $5m. It depends, it's all relative, based on the project. If I am making a machine learning infrastructure tool, nocode is a poor choice. But if you need to make a marketplace, job board, landing page I would just go for no code hands down. I just feel that my time is spent much wiser trying to get customers & validating asap.

    1. "Finally, can you build a scalable business with little investment without software involved?"

    No, you can't build a "scalable software business" without code. But you can validate demand and get first customers without code.

    Summary:
    Right tools for the right job.

    Bonus:
    I do support your comment about learning though. I learned how to code and it was inherently valuable in my sales job particularly when communicating with devs or how "internet infrastructure" works, helping with product management etc. Again, depends on what is the context.

    IMO:
    It just feels that lots of devs (no offence) feel bitter about the whole thing. #NoCode is not here to steal anything, rather give higher productivity & optionality power to both devs and non-devs.

    1. 2

      no offense but think you missed my point

      NoCode movement is not here to replace developers or take their jobs

      did i say this? no

      or anyhow imply that code is not necessary. It is here to make rapid prototypes faster.

      but it's wrong. you are not faster. yes if your use case is eg PH Ship but if you just add a bit more you are MUCH slower. Webflow hast a great UI but it's still clumsy AF if you can code. cumbersome web uis are good if you dont know what an editor or server is but you will stay slow.

      No, you can't build a "scalable software business" without code. But you can validate demand and get first customers without code.

      another fallacy, sometimes you need more to validate demand and a shitty no code landing page wont convert

      It just feels that lots of devs (no offence) feel bitter about the whole thing.

      utter bs and besides i am not a coder (and never have been). coding is imo a basic skill like speaking English everyone should know.

      1. 1

        did i say this? no
        fair
        but it's wrong. you are not faster. yes if your use case is eg PH Ship but if you just add a bit more you are MUCH slower. Webflow hast a great UI but it's still clumsy AF if you can code. cumbersome web uis are good if you dont know what an editor or server is but you will stay slow.
        no you are not, you are doing it wrong then. If Webflow is clumsy then I guess you won't be satisfied with the majority of the products on the market.

        another fallacy, sometimes you need more to validate demand and a shitty no code landing page wont convert
        what are you so bitter about really? Who talks about LPs? You can build a fully fledged marketplace now or a mobile app. Take a chill pill.

        utter bs and besides i am not a coder (and never have been). coding is imo a basic skill like speaking English everyone should know.
        it is not the first time someone is overly emotional about nocode movement on IH

        1. 1

          you started to use terms like 'BS', i just echoed your tonality. you think i am using nocode tools wrong? this comes from somebody who is not able to quote other posts properly[1]. I am a bit confused...

          [1] just reply to my comments without indenting yours otherwise your comments show up in a monospace font (like code)

          1. 1

            well, I no way meant to offend, but your post is really a statement which I strongly disagree with and the tonality was merely an emphasis.

            I don't think my honest mistake in "typing" reflects whatever I stated. It seems that you just confused the making of an LP with making the functionality of the MVP. Nevertheless, don't see a point of discussing it really as you frankly have a black & white opinion, I am just saying there are gray areas that you perhaps should explore.

    2. 1

      Agreed. I’m a developer and I’m using NoCode because I know how “expensive” it’d be to develop something on my own just for validation purposes vs do a NoCode solution that gets me there in 1/10th of the time. And I’m constantly surprised by the capabilities of NoCode solutions.

      1. 1

        thats the beauty of it, you can use whatever you feel is right for the job! I just hate "black" and "white" approach towards either something is useless, or not...

    3. 1

      If you think that all these tools have no learning curve then you are largely mistaken. They still require some knowledge of design (grids, paddings, containers, divs etc.), dataflows (merging tables, transforming data, JSON formatting etc.), understanding relational databases (CMS) etc.

      Do you think the "right tool for the job" might be dependent on how quickly you can pick up this sort of knowledge? I find that as I get faster at learning technical topics I'm more drawn to working at a lower level and often building things I once would have looked for a library to handle.

      1. 1

        I think its more about the type of product you are building. For something super simple you are good to go and use Carrd for instance. For more complex stuff there is definitely some custom code required.

  10. 3

    This comment was deleted 3 years ago.

    1. 2

      You couldn't have built the exact product that SongBox is, but maybe you could have built something similar enough and with enough of the features (except manually driven) to validate the idea?

      I am a developer myself but constantly trying to challenge my own assumption of "code first".

Trending on Indie Hackers
Why Most Startup Product Descriptions Fail (And How to Fix Yours) User Avatar 99 comments We just hit our first 35 users in week one of our beta User Avatar 44 comments From Ideas to a Content Factory: The Rise of SuperMaker AI User Avatar 27 comments AIgenerationtool — replacing hiring writers with 1 AI dashboard User Avatar 25 comments Why Early-Stage Founders Should Consider Skipping Prior Art Searches for Their Patent Applications User Avatar 20 comments NanoBanana or Seedream4.0? Why Choose When You Can Have Both User Avatar 20 comments