Creating a web-based app / platform yourself

[Lucky8]

I keep hearing stories of subject-experts/business people creating a web-based app/platform themselves, rather than pay an app developer a fortune to do it. Also I hear of no-code/easy code app development tools.

Has anyone out there done it and could share how they went about it please? Or do you have general know-how about how this can be done? I'm technically savvy, but not in these languages: where do I start?

 

[billybob99]

You have to ask yourself, what is the point of this story - usually they're selling a course on how to build a web app/subscription platform for next to nothing.

There are a ton of no-code tools out there like bubble.io - it all depends on what you're trying to build.

 

[mattk]

It depends what you're trying to do. No Code applications often replicate either simple forms or spreadsheet capabilities in an app. This is fine if that's the capability you need.

Tools like Wix and other site builders offer drag-and-drop websites, but they have limitations and are not very efficient when it comes to the code they output.

If you want a proper web-based application such as a forum, e-commerce and so on, the there are numerous off the shelf tools available, which can be configured by someone tech savvy, but don't require coding skills.

Finally, if you want to build something from scratch, then you will need software development skills. There is generally quite a steep learning curve, but once you have got yourself setup with the appropriate IDE (the tool you write the code in), database platform and other tools such as a web server, then there are tonnes of resources online you are learn from.

In reality, it depends what you are trying to achieve. I would be looking at the third option in the first instance every time. Why start from scratch when you can leverage something existing.

 

[John Hemming]

The underlying problem is that generally users want a good experience. If it is easy to produce a good experience then lots of people will have done it already and you have a marketing challenge. If it needs quite a bit of good coding then it is often better to get someone who has the experience. Happens that I do my own coding on my own service, but I have a colleague as well.

 

[ffox]

I keep hearing stories of subject-experts/business people creating a web-based app/platform themselves, rather than pay an app developer a fortune to do it. Also I hear of no-code/easy code app development tools.

Has anyone out there done it and could share how they went about it please? Or do you have general know-how about how this can be done? I'm technically savvy, but not in these languages: where do I start?

Depends greatly on what you need to do. Complex public facing apps can be very complicated. Simple internal usage apps to access your own resources for functions like stock check, customer contact lists and other resources can be achieved with a single click. Say more about what you want and you'll get a better quality response.

 

[gpietersz]

I think @mattk is right. There are lots of things out there. There are very good odds that something exists you can use with the right configuration, or just simple tweaks.

I keep hearing stories of subject-experts/business people creating a web-based app/platform themselves, rather than pay an app developer a fortune to do it.

Can you give us some examples?

As other have said you are not going to get meaningful answers without some indication of what you want to do.

 

[antropy]

I'm technically savvy, but not in these languages: where do I start?

Maybe in the first instance hire a freelancer to work alongside to create the app so you have their support and not doing it alone. Alex

 

[gpietersz]

Also, what do you mean by "technically savvy"? Options available to you depend on what skills you have so maybe if you can be more precise we can make suggestions of what might suit you.

 

[Paul Norman]

If you are sufficiently techy savvy, then this is fine.

My first question though, is whether it is the best use of your time. I find, in business, that if I stick to my strengths I can earn enough to pay other people to do the stuff that is not the best use of my time.

Obviously, I cannot answer that question for you.

 

[Lucky8]

OK good points. So by tech savvy, I mean I was a programmer many moons ago. I coded etc, but I didn't develop apps, whether android/iphone phone ones or the web apps I am talking about. As Paul's point says - it's not a great use of my time. BUT if I could find a way to combine my tech aptitude + my knowledge of UX, I think there's a chance I could put it together myself with a tool, with the areas where I really want to spend my time (on the business, the marketing etc) which is the best use of my time. Note - I have *no* intention of ever learning to code again, I am only interested in simple zero or little-coding tools. I want to create, get it out, see what the feedback is, evolve, typical Agile, typical fast prototyping, grow, get others on board to take over all the tech, I focus on the business.

As to what kind of web app it is (it may be accompanied by a phone app, but I'd rather not). In terms of functionality, it would be similar to the employee benefits and reward web apps/phone apps that are around. User profiles, database of information fed according to time/user profiles in a visually easy and appealing way, that frequency could be changed if it's annoying, ability for ad-hoc notices sent out by "super users" or my team. No need for any payments, there's nothing transactional going on here.

 

[ffox]

User profiles, database of information fed according to time/user profiles in a visually easy and appealing way, that frequency could be changed if it's annoying, ability for ad-hoc notices sent out by "super users" or my team.

Microsoft 365 and G-Suite from Google provide this. It's all web app from the ground up, and it can be accessed from a smart phone. No coding necessary at all.

 

[gpietersz]

I mean I was a programmer many moons ago.

That makes learning to develop it yourself a lot more doable. What languages do you know? What sort of work did you do? You might be able to find a language and web framework that fit well with what you already know.

I have *no* intention of ever learning to code again

In that case hire someone else to do it. Zero code and easy to use is going to be very limiting.

As to what kind of web app it is (it may be accompanied by a phone app, but I'd rather not). In terms of functionality, it would be similar to the employee benefits and reward web apps/phone apps that are around.

So the app would be the business? its an SaaS offering?

If it can be done with no code tools, does your business model take into account that potential customers could use the no code tools themselves to replace it?

 

[gpietersz]

Microsoft 365 and G-Suite from Google provide this.

Your invariable answer to everything.

You have not been working for the NHS recently, by any chance, have you?

 

[Lucky8]

That makes learning to develop it yourself a lot more doable. What languages do you know? What sort of work did you do? You might be able to find a language and web framework that fit well with what you already know.

I'm talking C++ etc, and for a short while I was a developer. This was a *very* long time ago, I wasn't good at it, and I haven't coded a single line of code since.

In that case hire someone else to do it. Zero code and easy to use is going to be very limiting.

Could I ask you and those who say there are ways to explain to me why you think the other person's wrong? In the middle here, and trying to get deeper info about why some think it can be done in the way I describe, and some think it can't. Why? Why not?

So the app would be the business? its an SaaS offering?

Yes, business model takes the DIY route into account. It's SaaS packed with information which they don't have.

 

[gpietersz]

I'm talking C++ etc, and for a short while I was a developer. This was a *very* long time ago, I wasn't good at it, and I haven't coded a single line of code since.

The languages you would use for web development are a lot nicer, and easier to learn than C++. On top of that web frameworks are also highly productive. You might be pleasantly surprised if you try it.

The more flexible no-code/low code platforms seem quite complex and require a learning curve themselves. It seems to be that "no-code" ultimately means "disguised code with a nice GUI" and you are going to need the same skills and a fair bit of knowledge to use them.

Some low code platforms admit that what they are doing is declarative programming. Simple things with some web frameworks come pretty close to this - for example if you add a subclass with just some string or number attributes and no methods.

In short, if you do anything beyond a certain level of complexity you will need to learn stuff anyway. You might find its just a nice GUI around doing the same work. Its not dissimilar to how a complex excel spreadsheet can end up requiring quite a lot of skill to do right. A visual programming language is probably even closer, albeit less familiar, example.

Of course if what you are doing is show some forms, fill in and store the data, and relatively straightforward browsing of data, no code might work well for you.

My experience is that most web apps, even ones that look simple, end up requiring coding some algo, or doing some complex database queries somewhere along the line

Could I ask you and those who say there are ways to explain to me why you think the other person's wrong? In the middle here, and trying to get deeper info about why some think it can be done in the way I describe, and some think it can't. Why? Why not?

It depends on what you want to do. You should probably take a look at the example sites, documentation and example code (or code substitute) for both no code/low code options and some web frameworks. Maybe bubble or similar for the former (they even have an app you can play with, without even registering). Maybe Ruby on rails, Django (my favourite) or Symfony for the frameworks. Then you can make up your own mind - and if both approaches are capable of doing the job, it might just come down to what you prefer.

Looking at the Bubble "showcase" sites most of them are very limited compared to similar sites built using code.

 

[billybob99]

This was a *very* long time ago, I wasn't good at it, and I haven't coded a single line of code since.

And so it should be, if you want to run an actual business vs being the limiting factor that tries to wear all the hats.

If building it yourself isn't an option, look at ways to validate the idea first. There are so many startups that marketed their idea as if they'd built it.

I've validated SaaS ideas by throwing a few hundred into UI mockups and hammering PPC.

From the sounds of it you shouldn't even be looking at building anything, bespoke, or no-code or anything in between.

 

[gpietersz]

I've validated SaaS ideas by throwing a few hundred into UI mockups and hammering PPC.

I think that is good advice.

From the sounds of it you shouldn't even be looking at building anything, bespoke, or no-code or anything in between.

Not yet, you mean?

 

[mattk]

I've validated SaaS ideas by throwing a few hundred into UI mockups and hammering PPC.

What level of conversion have you seen between people expressing an interest in your mock-ups and actually paying for a product?

 

[Chawton]

What level of conversion have you seen between people expressing an interest in your mock-ups and actually paying for a product?

You can test for this explicitly simply by deploying a mock up 'buy now' button. A courteous message thanking them for their interest and explaining that the final product is in development/being finessed will give you a very meaningful answer.

A simple UI mock up can give you a very solid answer on paid uptake (cf retention).

 

[Lucky8]

I've validated SaaS ideas by throwing a few hundred into UI mockups and hammering PPC.

From the sounds of it you shouldn't even be looking at building anything, bespoke, or no-code or anything in between.

OK useful to know, can you tell me more? How did you start, then what did you do with what? How did you enable PPC (pay per clicks?) without a fully functioning application and only UI mockups? Did you do the mockups yourself, if so what with?

 

[ffox]

This thread gets dafter. The OP said -

As to what kind of web app it is (it may be accompanied by a phone app, but I'd rather not). In terms of functionality, it would be similar to the employee benefits and reward web apps/phone apps that are around. User profiles, database of information fed according to time/user profiles in a visually easy and appealing way, that frequency could be changed if it's annoying, ability for ad-hoc notices sent out by "super users" or my team. No need for any payments, there's nothing transactional going on here.

M365 delivers the functionality of Lists. These are data storage containers or small database tables.
All the OP needs to do is decide what data he want to collect and display through a web app or mobile app,
He then adds columns to the List by naming a column and selecting the column type from a drop list. You do this through a web app and you view this through a web app.

Once he has populated the List he invites users to access it. This can be view only, or with editing rights.

If he want to restrict the access he can do this by adding users, or removing users, he can filter the list to display by date range, username, or any other key criteria he needs,

There is no coding and no technical skill, other than reading a set of instructions, required.

Dead simple.

 

[billybob99]

I think that is good advice.



Not yet, you mean?

My bad. Not yet.

 

[billybob99]

You can test for this explicitly simply by deploying a mock up 'buy now' button. A courteous message thanking them for their interest and explaining that the final product is in development/being finessed will give you a very meaningful answer.

A simple UI mock up can give you a very solid answer on paid uptake (cf retention).

This.

It also depends on the numbers.

I've tested an idea where I got 100 people to give me their email address so they could be notified when the SaaS launched.

To me 100 people for that particular idea was enough validation. Others I've aimed to get 500 people on the wait list.

Out of the 100, around 23 joined from the 1st email - it was a £99 a month SaaS.

2nd email, nobody.

3rd email, 4 more.

4th email 2 more.

 

[billybob99]

OK useful to know, can you tell me more? How did you start, then what did you do with what? How did you enable PPC (pay per clicks?) without a fully functioning application and only UI mockups? Did you do the mockups yourself, if so what with?

Its pretty simple nowadays.

I hop over to UpWork and tell my UI/UX guy, hit me up with a few landing page concepts for my genius SaaS - The Amazon FBA £1million a Month Product Finder.

He knocks up a few concepts.

I get it sliced up by a dev, hooked up to collect emails and join it up with a domain I have in mind.

I hammer it with PPC.

I determine if this idea has legs or not.

 

[Lucky8]

Dead simple.

To use the employee benefits rewards web app/apps analogy, I think there's a lot more to it than that. Not sure what you find so "daft" about this thread?

 

[billybob99]

M365 delivers the functionality of Lists. These are data storage containers or small database tables.
All the OP needs to do is decide what data he want to collect and display through a web app or mobile app,
He then adds columns to the List by naming a column and selecting the column type from a drop list. You do this through a web app and you view this through a web app.

Once he has populated the List he invites users to access it. This can be view only, or with editing rights.

If he want to restrict the access he can do this by adding users, or removing users, he can filter the list to display by date range, username, or any other key criteria he needs,

There is no coding and no technical skill, other than reading a set of instructions, required.

Dead simple.

You are a good guy, but Bill has got you seeing nothing but the Microsoft icon thru those thick glasses and that Microsoft start-up tune playing in your head.

There's a time and a place for using MS products to do stuff - and this ain't one of them.

 

[Lucky8]

I hammer it with PPC.

After you've hopped, been hit up, knocked, sliced, and hooked up -) ), where is your app residing for you to hammer it with PPC?

How many screens/how much navigation do you have at this stage? Do you have a dummy database serving dummy data to illustrate what the app does?

Can you give a rough costing for the interface design, developing the concept, and then once you believe you've proved the market, to develop the app fully?

 

[billybob99]

After you've hopped, been hit up, knocked, sliced, and hooked up -) ), where is your app residing for you to hammer it with PPC?

How many screens/how much navigation do you have at this stage? Do you have a dummy database serving dummy data to illustrate what the app does?

Can you give a rough costing for the interface design, developing the concept, and then once you believe you've proved the market, to develop the app fully?

There is no app, its a landing page with what it would look like, features, pricing table etc. I can gauge interest based on how many click on the pricing tab, how many click subscribe now (this page takes users to the sorry not ready yet page to enter their email).

Have max 3-4 pages. There is no dummy database whatsoever.

• Interface design if using folk from abroad, very talented lot in Serbia and whatnot. £75-£90.
• Slicing and dicing, £50.
• Domain, £10.
• Hosting, few quid.
• PPC, £100 max initially. Increase if getting good vibes.

To build out web apps fully, I've paid anywhere from $5k (USD) to $15k (USD).

You're gonna have to do all this marketing malarkey once you've built the thing, only to find out nobody is interested.

You might as well do it first and be sure there will be paying users at the end of the journey.

 

[Lucky8]

Thanks billybob99. That's all really useful.

 

[Lucky8]

Wait, I'd like to ask one more thing - from a tech perspective.
If I were to map out all the components and processes needed to code a piece of software that does this in C++ back in the day, I could.
But this is a different environment. Given the description I gave, what in your opinion are the components/processes/skills found on UpWork I need to pencil in and plan ahead for (after the concept is shown to have interest) to get a web app + phone app fully built?

 

[fisicx]

Stop calling it a web app. It clouds the issue.

All you need right now is a simple website that promotes the product and a form to sign up. A halfway decent developer could knock this up an a couple of hours.

Once you have tested the viability you can then look at doing some mapping. From what you describe it's just a database with some user facing forms to add the data and a bunch of reports. It's not a difficult thing to build. It's just a standard CMS website

Keep it really simple to start with then add functionality as needed.

 

[Lucky8]

Stop calling it a web app. It clouds the issue.

Not quite. I am distinctly referring to a web app, and not a website. They're different and being clear from the outset, far from clouding the issue, gives clarity.

All you need right now is a simple website that promotes the product and a form to sign up.

That's not it at all.

It's just a standard CMS website

Afraid not. The functionality I referred to requires interaction and dynamic elements.

Now, I'm happy to be shown how a website can achieve this functionality, but I don't think it can. So let me give you some clearer examples. Take 100 users, with 100 profiles with preferences as to the type of content that is served to them on their device (which may be desktop or phone). All the data in a variety of formats is stored somewhere. They choose the category of content they're going to get, this may change in 6 months, they need to be able to add more categories or stop. I splice that category into loads of different ways and formats, we both decide on the frequency of them being served some of the data (even when they're not actively in the app, like a calendar reminder), there may be stuff where they need to answer some questions, and their answers change what is served up next. etc etc.

All my research to date would indicate that is a web app.

 

[fisicx]

A CMS can do all that. It's just a bunch of DB queries.

That's the whole point of a CMS. It get relevant data from a database in response to a query.

 

[Lucky8]

A CMS can do all that. It's just a bunch of DB queries.

That's the whole point of a CMS. It get relevant data from a database in response to a query.

OK, so help me understand. If my 100 users need to answer questions and interact with content, and those different answers/reactions affect the next content delivered to them - as in which content is served up from the database, maybe its timing - a website using a CMS can do that? If I want to be able to allow my clients (not the users) to add their own content, can that happen too?
How would the CMS website look on a mobile phone compared to a web app?

Using the example I gave, what is the difference between developing a web app and a website with CMS from a cost, timeline, who-needs-to-do-what perspective?

And again using the example I gave of an employee rewards/benefits system, where employees can access and are served up content at their desk, on the move, without doing anything - are the ones out there built as websites using a CMS or as web apps?

 

[gpietersz]

A CMS can do all that. It's just a bunch of DB queries.

So are a lot of web appa.

You can reasonably classify a typical CMS as a web app. However not every web app is a CMS.

You can certainly extend or modify a CMS to provide custom functionality - but its no a no code approach unless the CMS already has the functionality you need.

Whether using a CMS is an efficient (particularly in terms of developer time) approach depends on what you are trying to do. If you are, for example, putting up pages of content and want to restrict access to particular pages to particular users customising a CMS to control access to particular pages would be a good approach.

On the other hand, if all your data is going to be sitting in custom database tables, and you are going to need custom code to edit (or import the data) and present the data the right way, then you are better off with a web framework which is designed for that use case.

You can, of course, use both, and depending on what you are doing you may or may not need to integrate them.

 

[fisicx]

OK, so help me understand. If my 100 users need to answer questions and interact with content, and those different answers/reactions affect the next content delivered to them - as in which content is served up from the database, maybe its timing - a website using a CMS can do that? If I want to be able to allow my clients (not the users) to add their own content, can that happen too?

Yes

How would the CMS website look on a mobile phone compared to a web app?

I will look however you want it to look - it's just CSS

Using the example I gave, what is the difference between developing a web app and a website with CMS from a cost, timeline, who-needs-to-do-what perspective?

No idea without seeing the spec

And again using the example I gave of an employee rewards/benefits system, where employees can access and are served up content at their desk, on the move, without doing anything - are the ones out there built as websites using a CMS or as web apps?

Both.

Rather than get hung up on the method of delivery, map the workflow and put it out to tender. The responses you get will have a range of solutions. You just need to evaluate the best one for your business idea.

As to delivering content depending on the question asked, this is something I built for a client in the USA. What you choose in each question determines the options in the next. It's just a simple DB query: https://variants.loanpaymentplugin.com/title-loan-application/

Or consider eBay when you make a search then apply filters to refine the results.

When ever you login to something on 'tinternet you are connecting to a server somewhere. You can call it a website, webmail, web app, SAAS or anything else. The name is not important, it's the functionality that matters. This is what you need to do. Map out the customer journey then sort out the best way to deliver this.

 

[Lucky8]

Map out the customer journey then sort out the best way to deliver this.

A little chicken and egg here, but I've generally done that. The benefits/rewards example I gave is fairly accurate in functionality, which is why I asked "what is the difference between developing a web app and a website with CMS from a cost, timeline, who-needs-to-do-what perspective?" for it. You say you can't answer that without seeing the spec, so beyond all that I've explained, what more in terms of spec would you need to answer that question? I'll answer it if I can.

Putting it out to tender is of course one solution. But the starting point of this thread is that I am seeing if I can do it myself.

 

[fisicx]

The full spec would include the whole customer journey. It would include all the form fields required for registration and login. It would detail what the client sees on each screen and how they move backwards and forwards. It would show the required queries and filters and how they get the data.

A spec for a project like this could run into dozens of pages. The simple example I linked to cost $600 to create after two weeks of detailed discussion. The actual coding took about a day to do. Another recent project had a 40 page specification and cost €1500.

The full spec is key to everything as it will define everything - every single option needs to be detailed.

A web app or website uses HTML and CSS for the page structure. It will often use PHP and JS to manage the data. Alternatives are ASP or NET.

If you want an App you download from the App Store or Google Play then it's a whole different ball game. These use Swift or Java/

 

[Lucky8]

So fisicx, I know about the importance of mapping, that's always been the first step no matter what language or tech anyone's using, and I'm already on the case.

Let's take your title loan application example. The user selects from a number of dropdown fields, and they're taking through subsequent selection fields depending on what they selected in the previous question, then they're asked to fill in their details.

If.. their user profile was set up already. They could a) login and view loan content on their desktop b) not log in actively but the webapp/app serves up information on their phone a few times a day, let's say about changing interest rates, or loan providers, or car tips. Maybe the desktop view aggregates all the "tips" they received for the day in a different format. They might get a tip sent to their phone which asks them "when was the last time you had your car serviced?" and they have to input a date, or they are given a little quiz which they have to answer. This interaction and two-way dynamic elements are aspects I was under the impression a website/CSS could not handle.

You and @gpietersz are giving me great tips here, but I'm still a bit cloudy between you both on the advantages/disadvantages of your different approaches and which way I might need to go.

 

[fisicx]

You are describing two different animals. A webapp or website can push notifications to the device if permission is given. But what you describe isn't really a single application. It's more like a website for the desktop/tablet user and an app for phones. And as I described above apps are whole different thing.

Two way dynamics is just a matter of development. You can do whatever you want with whatever you want. It's just code.

At the monent we are just guessing as we have no idea what you want to do.

You mention profiles - these are easy to create as it's just a form. The data is saved to the DB and when they login this data is used to deliver the appropriate content. Nothing complicated at all. If they change their profile when the DB query is made a new set of content is delivered. This is how a CMS works. But as @gpietersz suggests, without knowing the detail nobody can say which platorm is going be most suitable. You may even need a combination of things.

Either way, it's going to need coding. Even building a simple registration form with validation and processing needs coding. You can't just drag and drop something into place.