Full disclosure, I'm a software engineer. And the recent trend of no-code solutions have been keeping me awake at night.
Do you think custom web/mobile development as we know it today may die out in the next years? Is everyone going to be drag and dropping stuff to make their products work?
Would love to hear your take on this, is it going to pass or am rightfully I paranoid?
UPDATE: Just to keep the discussing on the intended path - This is a thread about what's to come and speculating about the near future, rather than a "will robots replace us" type of thing :)
I see no-code as a great way to validate ideas quickly. Many lightweight browser tools and creator utilities can launch much faster this way.
For example, projects like SSYouTube Download show how browser-based utility tools can solve simple user problems efficiently. In niches like video downloading and conversion, speed-to-market matters a lot.
That said, once scale, performance optimization, and advanced features become important, custom development becomes hard to avoid.
So in my view, no-code is excellent for early growth and validation, while full development is better for long-term scalability.
I think no-code has lowered the barrier to building useful products, especially simple web utilities that solve one clear problem. Users care more about speed, reliability, and ease of use than the technology behind the product. A good example is browser-based tools like igram, which let people download publicly available Instagram Reels, videos, photos, and stories without installing software. Whether a product is built with no-code or traditional development, the real measure of success is the value it delivers to users.
The way I look at it is no-code is like Legos. You can build anything as long as you are happy with the limitation of Legos.
This means that it's not optimized or robust or customized the way it would be if you custom built whatever without Legos. Now if you know Legos you know there are folks that can build really amazing things and those that can barely put 2 of them together. Just because you have a bunch of Legos doesn't mean you can create with them.
I think it's the same thing here with no-code. It fits the mental model of more non-coders but not everyone. If you go look at any of the no-code forums you'll see many people posting jobs to make something using a no-code tool. So obviously they don't want to learn to drag and drop - just accomplish their business goal cost effectively.
So I think you should be mindful of what's going on there and have a plan if the trend accelerates faster than is anticipated. If it doesn't then you can continue until the next disruptive tech shows up in a year or 2 ;)
I built Ministry of Testing on no-code for a good 7/8 years before moving over to 'real code'. We got to around £400/500k annual revenue before committing to 'code'.
I started in 2007, the tools I used seem pathetic compared to what is available now. However, we still use no-code tools (Wistia, Mailchimp, Intercom, Discourse), but our code connects with these things with how we choose and in a more integrated way.
Currently how things are, we definitely move slower now with how and what we build. Partly because of lack of dev time, but also (I guess) we are a bigger team. It's at about $1.3m annual revenue now, a mixture of events and online membership.
I did have huge frustrations at the time. I tried building membership and recurring memberships with WP a few years back, I hated it, especially the lack of control of user flow/experience.
But also on the other hand, I realise I was very much focused on getting stuff out the door and seeing if people wanted it. I rarely spent too long on an idea before checking/seeing if it worked.
I'm personally excited about the no-code movement atm, especilally as a non coder. My husband appreciates not being asked so much to build stuff as well! I've been dabbling in a few no-code things, but I really wish I had more time to explore it.
But I also can't see why there wouldn't be a need for dev skills. Things are just getting more complex and it's hard to build anything from scratch. Integrating with other things will become more normalised?
It doesn't seem that developers will run out of jobs anytime soon. The complexity of applications increased significantly in the last decade, and it will continue to do so. As we understand common application patterns some apps will be developed by no-code tools regularly and typical. That is actually good, so you can concentrate your energy on the exciting stuff. This is actually, the goal of programming: increased abstraction and better access to solutions to non-programmers. I embrace this trend.
Pretty much sums up my thoughts! (in a much more succinct way 😂)
I've wondered the same thing. My perspective is as someone who tried desperately to be a developer for years but could never stick to it/wasn't clever enough/what have you. I couldn't get past HTML/CSS.
My general take: we will always need tech go-betweens. No matter your age, there are some people who just aren't "tech-minded". For example.. I know a LOT of people who can't even begin to fathom setting up a Squarespace site. To me, it was the easiest thing in the world. But to someone unfamiliar with coding or design, it's a foreign language.
I have been playing around with Bubble and Sheet2Site and find them fascinating!! Although this allows me to design and build without code, it still requires me to read instructions and watch tutorials... which is still time-consuming and risky (what if I don't "get it"? What if these platforms close one day?)
I think there are niches in the market that will shrink or disappear... but in general we will always needs technology translators who can bring ideas to life for the rest of humanity.
First, I’d like to posit that no-code isn’t new. We’ve been copy/pasting plugins into WordPress for years and creating fully functional apps without a line of code.
It’s becoming easier, though, to develop more and more complex apps without a lot of dev know-how. I should know. I suck at programming but still manage to hack together way more projects than I should.
Programming is definitely becoming commodified. But there will always be a place for more advanced users - applications which are too rare to build no code solutions, or too complex, or performance reliant, etc.
Being a “webmaster” was super lucrative in 2005...now, not so much. Everything evolves!
Indeed the idea is not new, I was referring to the shift where some services get packaged into products and become more or less obsolete - e.g. landing page builders starting to serve the original demand which actually drove a large chunk of today's web developers actually going into coding (small business websites, promotional marketing projects, etc). Things will keep moving.
I agree. Back in my web design days (before 2010) I used to charge $5k+ for a simple static web design build - and that was under market rate. Now, that's impossible. It also used to take me a week or more, now I could spin up the equivalent site in a couple hours.
Others hit the nail on the head - the abstraction of complicated technology such as code isn't new, and of course more and more things will be able to be accomplished without code. This doesn't at all translate to programmers being rendered useless (maybe until we have AI that can write all that code :)).
Consider how C programmers felt when python became popular. They probably felt the exact same way you did when more and more things were able to be done by less complex code in python. Python was not a no code solution, but "less code", and no code is another step in that direction. You'll notice that C programmers still exist and fulfill important roles, though more and more people will be attracted to the "easier" solution. Maybe in 20 years there will be very few programmers because they have mostly moved on to become "no code makers" but technical solutions themselves will always be needed.
I would like to mention that I think a lot of the nervousness over no coders taking over all the traditional software developers jobs is definitely magnified within the maker community (looking at the demographic of course it makes sense). You tend to see a lot of very enthusiastic and confident people talk about the power of these no code tools and rightfully so; becoming tech-literate and being able to create amazing procts/services without the traditional barriers of a cs degree or being able to code is extremely empowering.
The downside is for some reason this seems to lead to thinking like misinterpreting CS degrees as equivalent to html/css/js tutorials, or the over exaggeration of what no code tools can do compared to code.
Smart phone cameras didn't eliminate photographers and medium didn't eliminate journalists. Just like no code tools, they lower the barrier to entry for the average person, and I think that's a good thing.
This sums up pretty well my thoughts. As much as I like the idea as a non coder. And as much as I'm fascinated by all new solutions coming along for non coders, I really don't see a world where engineers would get obsolete.
what does next years mean to you?
there will always be more layers of abstraction making things easier and accessible to more people, low/no code is just the next step. Eventually we might reach a point where you can design a website with your voice alone.
Personally I'm not worried at all about my future in the industry and I'm not even 30 yet, but it will probably less demand in 20 something years.
Hi Yasen, I run www.skaffolder.com I strongly believe that developers will be still a must-have in the next 20-30 years.
This is my vision, so we built a code generator that is close to a no-code tool but still allows developers to have full control on the code.
My idea is to support developers, not replace them (us), opening the creativity of developers and let us focus on the code that matters (business logic) and not on the boring part of the code (scaffolding).
I remember attending Future of Web Apps conference in 2009 where one of the talks demoed a no-code tool to build a Twitter client (albeit the end result might have been a Flash application?). Already then we wondered if no-code was going to ever be a hit because it had been tried for a decade before that.
So the no-code "trend" has been around for at least two decades. I think it will likely remain a second, smaller track that can be useful for some purposes. If it allows people to build something useful then why not? The market for coding stuff doesn't seem to be shrinking, quite the opposite.
No code smells like kool-aid to me. I’m probably wrong, I often am.
That said, I sometimes think about the stability and portability aspect of no code solutions. If you build a solution in no code service ABC what happens to your biz when ABC changes its offer (api, cost structure, goes out of business, etc) you’re gunna have a bad time.
I do realise this might look like a similar problem if you rely on AWS (as an example) infrastructure when you create “on code”(tm) solutions, but the fact is aws is just built on tech that’s already available - it’s a convenience more than anything, and devs are well versed in responding to fundamental changes to things on which there solutions are built. Can the same be said for “no code monkeys” that can be said for the standard “code monkey”, I doubt it.
In the past I’ve built on PAAS services that have changed their offer, gone broke, been acquired or whatever and it’s a royal PITA to rebuild everything.
For anything remotely serious I’d never choose a no code solution that can’t be replaced within the time span of a week or less. And even cobbled together solutions from various vendors I’m suspicious of. Eg Auth0
There are a few significantly successful makers, I won’t name names, who build using basic building blocks like php, HTML, css and vanilla JS. Some work to replace everything from Slack to Google maps to ensure they’re master of their domain. I applaud this, 💯
These days there are so many resources to learn the basics, listed above, there is very little excuse not to. I don’t want to invoke the “when I was a boy” line, but I just did, there was f*ck all. Learning C was really hard! Man pages, and if you could afford it, 100$ or more books were your guides - $100 was a lot in the 90s. Now I can pull up a million well written tutorials, get the answer to almost anything you can think of in an instant, via google and stackoverflow etc.
I dislike the no code “movement” but not because I feel threatened, but because it’s like building your house out of sticks and mud on a base of quicksand. 🥳
Finally, as a customer I’d be skeptical of the longevity of anything built from “no code”.
This comment was deleted 6 years ago
Semi related side note: I’m interested to know what your thoughts are about ML, DNN with regard to this type of thing. How do you see this playing out? Will it be a few large vendors Azure, AWS, Clarifai, doing 80% of the heavy lifting in the app market or is this a wide-open frontier?
What does no code AI look like in the next few years?
I’ve been interested in learning “how to AI” 😂 for a couple of years now, but honestly I don’t grok the math well enough and so it’s been something I’ve been watching from the side rather than “getting in the pool”.
This comment was deleted 6 years ago
Thanks for the response.
That makes sense.
You make a good point, no doubt.
I was referring to the ecosystem fragility rather than quality of software or operations for that matter. But I don’t disagree that there are benefits to outsourcing some things. For example, I see using RDS as a benefit over running a DB on my own server. I see EC2 as a better option that sticking a server in a local data centre.
The difference between that and what I was referring to as “no code” is that i can easily move to an alternative if I’m using the fundamental and portable building blocks.
My discomfort sits with the lack of portability, lock-in situation as well as the higher level modularity, generally speaking.
It’s an interesting situation that I’ve been grappling with myself, in a variety of contexts, which can be boiled down to something like “where do I want the complexity to sit” and the answer will probably be dependent on skills, knowledge, experience and confidence with the situation under consideration.
This comment was deleted 6 years ago
I mean if someone doesn’t learn how to build software they’re at the mercy of what’s available. This is the core problem I have with it.
Perhaps my concern is not warranted.
It could just be a case of old dog new tricks.
As I said up front, I’m often wrong. :)
Update: Actually it’s not even a case of right or wrong. This is just how it looks from my perspective.
This comment was deleted 6 years ago
I got started in web development in 1997 because I could drag and drop blocks of design into my site on homestead.com.
It's 2019 and we've evolved the web a lot since then and drag and drop WYSIWYG editors are still popular but not putting actual engineers out of business.
This comment was deleted 6 years ago
I hope so, and I hope it's my work that is causing my work to be redundant. I'm not here to be stagnate and have job security. I'm here to do great things.
Hey! I'm interested in sharing my own experience in building with No Code
I am the founder of ConductScience.com - a company that makes translational science tools for outcomes in science.
Our divisions include digital health, hardware neuroscience mazes, virtual reality, drones, knowledge transfer, and any useful scientific tools that help scientists measure outcomes in science.
First and foremost, I am an emergency medicine doctor. I’ve spent over a decade training in medicine, so this means that I don’t really have much experience with coding. I can create websites using popular tools, but definitely can’t do anything more than soft coding.
When creating ConductScience, I knew that I should play to my strengths and stay away from coding as much as possible. Where needed, it makes much more sense to hand off complex but necessary programming jobs to people who are qualified to do them.
What platforms and tools have you used to build ConductScience? How long did it take to build?
I first purchased the domain in about 2012, with a pretty clear vision of what I wanted
I realized that I spent the majority of my time on unproductive tasks such as building maze prototypes and little time collecting data. I thought about how great it would be if there was a service from which I could buy ready-made scientific tools that actually measured outcomes instead of mechanisms. Every company had really complex tools to find out exactly how something worked without ever figuring out if things ever actually did work or not.ConductScience to become. The first realization was that we would need our own proprietary tools, so it took me a good three or four years to create MazeEngineers.com. I wanted ConductScience to have one true proprietary tool before expanding rapidly as we are now.
ConductScience is based on WordPress, so no coding was needed. I hired specialists to help me with some custom CSS for the site. In case you aren’t familiar with CSS, it basically describes the presentation of a website in HTML format. It isn’t a programming language, so there was no need for any coding there either.
How have you gone about growing your user base? What has been most effective?
Growing my user base and building a loyal following of scientists and researchers has come about, I believe, from focusing on making the product solve the pain points I discovered as a lab technician and scientist myself.
Almost a decade later, we have a much broader array of scientific tools which has made ConductScience become much more easily marketable. The tools allow scientists to spend more time conducting the studies that are at the core of what it means to be a scientist.
One step we took to help make ourselves more attractive to potential customers was to try and establish an ‘easy-access IP’ style of tech transfer for these tools. Results-based tools which measure medical outcomes are not as well patented as other mechanism-based tools, like lab equipment. The latter has a very organized model when it comes to the movement of products.
We made the effort to negotiate deals with the universities or manufacturers that develop tools we want to sell, and gain licenses to spread the technology amongst the scientific community.
On another note, we also use tools like Google Analytics, Ahrefs, and Crazyegg to better understand our customer behavior in the maze that is our website. This helps us to understand what actions we could take to maximize growth.
Have you monetized it yet and if so, what is your revenue model?
Yes, we are monetized! Importantly though, we are a bootstrapped business. This means that we grow with our own revenue, and never accept outside funding. I believe this is a crucial part of our revenue model, because accepting investment could lead us to compromise on the quality of our scientific tools by rushing to meet an investor’s targets.
In order to make revenue, we have to provide real value to customers. We work hard with all of our divisions to ensure our products are always going to stand up in a real scientific experiment. Put simply, if our customers are pleased with the tools they receive from us, we will manage to keep growing our revenue stream.
What has been the hardest part of building, growing or monetising ConductScience?
Making the decision to be bootstrapped is without doubt the toughest thing about growing ConductScience, despite the benefits it brings. Relying on our own revenue to grow means that we have to work really hard to cut unnecessary spending, and only work on things which have a great chance of a healthy return on investment. This discipline is both the hardest thing we do but also the most important source of our strength--it means we have little waste and everyone has the same expectations of productivity.
Do you have any tips on building, growing or monetising a no code product/business?
Yes - even though there is no coding involved in the process, don’t make the mistake of thinking that a functioning website is easy to architect!
I do liken the science of website design to that of building mazes for neuroscience purposes. Everyone who visits the site needs to land on the part of the site relevant to their needs, and it is our job to guide them there. Just like a maze, we need to construct the architecture so that each user finds the ‘cheese’ at the end of the trail.
Here are several tips that spring to mind when thinking about how to design a great no-code website:
You should think of yourself as the chief growth officer.
Knowing the product and customer base, how do you design websites that speak to their needs?
How do you design websites that deliver knowledge when they search? This is value that builds trust, eventually leading to revenue for you.
When you have revenue, who do you hire to grow the product?
The website should be a central growth machine to generate more revenue for the company, and all marketing activities should lead to funnel towards that.
If your site is offering skincare products, for example, make it easy for your customers to find their ‘cheese’ (their desired product). With an extensive stock and intuitive search function, you’ll already have gone a long way to designing a website that navigates clearly.
I know my strength lies in science and medicine, and not in coding. I think that the best thing you can do to grow a product like this is to take your strength as far as you can. For me, this was designing scientific products and conducting experiments to test their efficiency.
Where can we find out more about ConductScience and connect with you?
Sure - please feel free to contact us if you’d like to find out more about the products we offer. Our experts are on hand to chat, and if you’re a scientist, we’d love to hear about the things you’re creating.
Alternatively, if you really like the idea behind ConductScience and would want to help out in some capacity, please feel free to browse https://conductscience.com/work/.
---
Short bio: ConductScience is a company that specializes in the development and spread of scientific research tools across the community. These tools are often based on collecting results, like animal behaviour equipment, digital health and symptom trackers, or easier knowledge transfer
Hello @BetaPeak,
This topic so feels interested in me.
I think the future of web development will easier for Developers so much.
And the end-users who don't know any code can make a website in a few how included edit content.
Our team also working on one Website Builder that supported for WordPress and Shopify.
User will use our bigs library with full site layout included and ton of Blocks to select to build their brand.
This is our demo to see how it look likes: https://library.layouthub.com
We are in the last step to release it for Shopify first in this month then WordPress version.
if you are interested in just let me know when release I will send you a link to try.
Best Regards
How do you handle authentication in no-code app?
You can build a serious BUSINESS with no-code, but you can't build a serious tech PRODUCT.
At the end of the day, no-code tools were built with code.
I'm not a dev and this is how I feel about it.
I've not used no code solutions myself but I don't think it will replace actual coding. Although you can do a lot with these no code solutions usually with these drag/drop building block systems there's some limitation somewhere where the only way to get round it would to custom develop something with actual code.
Personally, I see software as a solution to a problem. When a problem is solved, we move on and solve the next. Software engineers are still in high demand in the next 10 years.
No-code is not a new thing. I built my first website in Dreamweaver in 2002. No-code/visual tools have been always there, they're just getting better.
Behind all no-code tools are still human engineers. I know some developers make half a million a year building Shopify plugins. New markets create new jobs.
It will never die imo.
A little bit. But the more I've thought about it, it seems like CRUD apps are the only things that will go full no-code anytime soon.
If you haven't checked https://bubble.io/ I encourage you to do so, there are actually some decent and pretty complex stuff built on that thing. I'm not advocating for it, just something that has been on my radar for a while.
I second Bubble. Webflow seems to get a lot of attention on twitter, I think it's maybe because they hire a lot of people active in the community. But I think Bubble is the better tool. Put it this way, I've been coding for about 10 years, and the last time I was blown away this much by what Bubble can do, is when iPhone apps became a thing.
I'm not concerned about the no-code solutions. They can only be useful to an extent, which isn't a bad thing because honestly maybe that's all some people need for their ideas.
Maybe those that work for full-stack developer agencies that are used to churning out MVP-level products for a low cost may be losing some sleep at night, but I believe the rest of us are fine.
@caitlinjohnson MVPs don't necessarily mean simple implementations put together in a hurry. Depending on the market you may need to put quite a lot of effort (and manage associated complexity) until you reach a point you can consider calling a product an MVP. And I should know because I own such a company, building such products :)
Even if people start using no-code solutions for MVPs.. then what. You get to a product/market fit and want to build more complex stuff on top of the functionality you have but can't because you're in a closed/non-extendable no-code ecosystem. So do you rewrite your system (terrible idea) or do you stay at the no-code platform, limited by its capability?
The thread was more about what's to come and where it's going, rather than a "will we lose our jobs" type of thing, sorry if it came out that way :)
Ah, the bit where you said "I'm a software engineer. And the recent trend of no-code solutions have been keeping me awake at night" gave me the impression that this was a "will we lose our jobs" post. But thanks for the clarification!
And yes, I agree that MVPs aren't always simple implementations put together in a hurry, and apologies if that's what my post indicated which was not my intention. I am also building such MVP as well, but even then, there are no-code solutions that can put together complicated ones (companies like Zapier and Makerpad make this very doable).
If everyone ends up using no-code solutions, then perhaps it would make it more difficult to distinguish yourself from others? Sure, there's customer support and branding that can help give yourself an "advantage", but I believe that some of the success that companies see derive from customized solutions that allow them to separate themselves from others, which may be more difficult to achieve with no-code out-of-the-box solutions. Not sure if that made sense.
This comment was deleted 6 years ago