How much does it cost to develop Clubhouse-like application?

[imaclarenceK]

Want to order a clone of Clubhouse from app development company. How much it would cost ?

[SwissProgrammer]

This is my opinion.

If you want to own it all, then:
At least one web server.
Hardware price depends upon you; From used (20+ years old) to newer. From free to a lot more. A MS Windows 98 Second Edition with C on a 286 SX could handle this, rather slowly, but it could handle it. If you want this on a not so old computer on a budget, remember that MS Windows NT 4 was (by at least one company), it seems, proven UN-hackable. You might not want to do this on any computer or server past MS Windows 7 sp1 unless you are willing to share a hugh chunk of your bandwidth with Microsoft's spying on your activities a lot. Do not use any version of Unix or Linux for this. Do not use any Macintosh for this. Sun Microsystems or up to Windows 7 sp1 might be nice for you.


Program language: C or up to and including C++20. Can be done with some others, but that is up to you.

If you are coding for yourself: Fast communications for this can be via a ping like package. Not ping, but ping like. If you do not know how to do that then study File Transfer Protocols.


If you are paying someone else to write it for you: I expect a high quality coder might be able to complete this in about 1 to 3 weeks at 5 days per week and 8 hours per day.

To get ahead of the Clubhouse app people: Make your project a GUI with a CLI option, and make it Android, iOS, iPhone, Windows, Macintosh, XBox, and PlayStation capable. Make your program's XML in two versions (for separate executables) one that is app store enabled and one that is specifically NOT app store enabled. I wrote an article that is on bytes.com on the use of an XML with C++.

Cost to code: Maybe around $9,000 to 30,000. Tell the programmer that you own the raw code exclusively. Tell them to NOT use Visual Basic, or Visual C++, or .net, or Active X for anything in the program. Tell them to NOT use a Microsoft brand compiler, and to NOT use an Intel brand compiler. Use a GCC compiler for this. Do not use a Qt compiler for this. Get bids from at least 5 programmers that you meet in person before asking them to bid. Do NOT contract or even accept bids from anyone that you contact over the internet or that contact you over the internet.

That is it.

You ask, now you know.

[julzyw]

Good question! I'm also interested in this topic.

[Niheel]

Swiss,

This caught my attention

You might not want to do this on any computer or server past MS Windows 7 sp1 unless you are willing to share a hugh chunk of your bandwidth with Microsoft's spying on your activities a lot. Do not use any version of Unix or Linux for this.

. . .

Use a GCC compiler for this

Why would one not use Unix or Linux? Also Windows 7 EOL was January 2020?

[SwissProgrammer]

I was speaking of producing a commercial product for use on various systems, addressing systems to develop it on.

I currently consider that Unix and Linux are not sufficiently standardized for my comfort with certain system to system communications. They are standardized, but they can vary in this area. Such as, it seems that many banks and other need-to-be-secure systems, if they use Unix or Linux, then have a version of those operating systems that are only compatible via exterior or surface communications, meaning not compatible via file to file communications. I am not saying all, but many, by design for security. A ping-level type of two-way communication might not work well with them. Unix is a powerful operating system that can be very secure, but there can be many flavors of Unix that are designed to not be compatible at this level of communication (specifically for security), so for a commercial universal program I suggest not using Unix.

For this type of a commercial program. Not a universal opinion to be applied for all other programs written. Just for a system on which to develop a clubhouse-like application.

OK, that might not explain it enough, so I will give some background that might help.

Linux has literally cost me. I have found it so full of hack-able holes that I do not want anything to do with it other than to suggest that other people not use it for anything.

As a separate outside example: A very big and very powerful and (for the internet) very old system company is at networksolutions.com . I use them even today. I have no stock in them. The told me that they prefer to use Linux. I questioned their logic, but still used them. Then that big, powerful, stable, company that hosted multitudes of web sites was hacked hard and I read that maybe a million or more of their customer's information was stolen. I am still there, but I am not their average customer. I would prefer that they used XP 64 with limited selected (by me) service packs or 7 64 with no more than sp1. But, I am dealing with it.

EOL or even EOS from Microsoft. Ha! Any L or any S from Microsoft. Ha! I do not know of any customer anywhere that I am clear that actually received any actual help from Microsoft (as it related to a customer request for help) ever. Ever. Ever. Ever.

Not only that but, back in about 1996 when I was a President of an Internet Service Provider system building corporation, that built and configured and installed the network systems for ISPs by contract, when we found a hackable hole in a Microsoft Windows operating system, and when we asked Microsoft for help, they not only did not help us but also did not help the Microsoft "Most Valuable?" Professional that we had on staff and whom was a stockholder in our corporation. No help. We came up with the patch for the operating system ourselves and "gave" it to Microsoft because we felt sorry for Microsoft being so inept. EOL? There was not a beginning of it. It is a void being advertised and pushed hard as though it exists. But I have never met any one that experienced even it's beginning. It is like smoke and mirrors, and advertised a lot.

I have, since then, researched and tested Microsoft operating systems and found the most stable to be, in this order NT4, 2000, XP, 2003 server and last and maybe least 7. From Windows 8 to and including Windows 10; I found to be unstable. Windows 10 I found to be the most unstable and the most full of security holes. Windows 8 I found to have less Microsoft back-doors built in than Windows 10.

Direct experience and actual cost has pushed me to not use any Linux or any Microsoft Windows operating system past XP sp2 or 7 sp1 or Server 2003, all of which I am comfortable with.

Back then, AT&T used to send problem customers to us that they could not or did not feel like handling and we made money building systems that worked. We had a T1 to our office, and later, after I shut down the corporation, I found that AT&T had an extra T1 set aside just for us in case we might want it. That was with us using Microsoft operating systems. Some systems work well. Some systems do not work so well. It was our business to know which major operating systems did and which did not.

I am being careful not to say the name of the corporation, so I can tell you, when my techs would build a system, I would have them test the system on-line in our facility for a while. While they tested, I actively (without telling them) posted challenges in newgroups and to hacker groups begging them to penetrate and crash the system. I did not allow the systems to be released until the hackers and perpetrators were stopped. Sometimes we had to come up with the operating system fixes in-house and had to fix Microsoft's screw-ups. But, we did it.

I was not a C or C++ programmer back then, but our corporation had quality personnel and we had success. Back then, some of what I learned about security and networking, I learned from them. Now I am attempting to learn C++. I have also researched and studied various systems to develop commercial programs on via C and C++ and have found the list as I have described in this posting.

I hope that made sense. I hope that I was not too harsh, but rather starkly honest.

Microsoft is just a company trying to make money for it's stockholders; nothing more, nothing less, and some of their products are really nice and some are really not nice. I have experienced Microsoft's customer service, multiple times, and other than their automated response system, it seems to actually not exist. No BOL. No L. No EOL. They send out a lot of inept adjustments to their currently inept and very hackable and very back-door pre-infiltrated systems. They can (to a very small extent) claim "new." New is not by default better. Other is not by default better. In Microsoft's case, their operating systems are getting worse.

Thank you for asking.

I like Microsoft (some of their products).

I think that Unix can be very secure for banks and financial institutions (but not a universal platform to develop commercial programs on).

I do not intend to use Linux for a can opener.

[Niheel]

Swiss,
At first, I thought you were trolling the OP with the recommendation.

Thanks for providing more context for the approach you presented. It's appreciated.

I was speaking of producing a commercial product for use on various systems, addressing systems to develop it on.

On the other side of spectrum, the clubhouse service is built for mass communication as a platform.

I would imagine clubhouse stack is a cloud back-end. Most likely AWS/GCP with the iOS front end ( which was probably developed on a mac using xCode ). There's a little bit of information out there that they use/used Agora (API) as the provider for their voice communication.

Building for mass market requires more speed in development to accommodate for the experimentation and there's a greater deal of linux based products and services to build upon than there are for other operating systems. It comes down to the use case of the end user and the trade-offs to allow for developer speed and flexibility.

Linux is everywhere. Even on Azure, more than 50% of the VMs are linux based. Some of the back-end of Azure like it's SDN are linux based.

One of the neat aspects of software development is the diverse approaches to solutions and also the diversity of use-cases for the same technology.

Thanks for sharing yours experience Swiss.

As to the OP's question on how much it would cost:

It really depends on the use-case. If you are building for mass-market like Clubhouse. They've raised $100M+ to build, grow and maintain their service. A lot of that would be engineering and product costs.

Building the service as feature on an existing internal or external platform would be much cheaper but still expensive. The cost has an inverse relation to the level of internal development talent on your team.

Twitter and Telegram are two products that are building similar add-on services. I've seen smaller companies do similar add-ons. I would imagine the smaller companies spent no more that $50,000-$100,000.00 to make that happen.

[TatsianaIsakova]

My team has recently shared their take on Clubhouse-like apps. Although the article doesn't include an exact number, you can read on Clubhouse tech stack and in general find out more about its business model.

As for the costs, is you are aiming for a basic Clubhouse-like app, it may cost you somewhere between $20k to $30k. However, if you'd like to develop a more advanced version, the price is increasing in proportion to the number of features.

[SwissProgrammer]

TatsianaIsakova, if your team programmed a Clubhouse-like app, and I am not certain whether you are saying it did or did not, then you probably have some idea of what it took your programmers to do that.

Logic base for the following:
It is common in the structural steel industry to use a "take off" to start construction estimates.
In the home building industry that might be called a materials estimate.
For a secretarial pool, an estimate for the production of a document might start with a word-per-minute speed estimate of the secretary(s) doing the typing.

So, an estimate should be based upon some known logic. Not just a wild guess. Lets go with that:

My typing was tested and found to be about 55 to 60 words per minute including fixing spelling mistakes. I have approached maybe about 80 or 90 wpm sometimes, but 55 is OK. So, let's use a conservative of about 50 words per minute for this. If someone types 50 words per minute and types for one hour, non-stop, that is 3,000 words. If they do this for 8 hours a day, with enough breaks that the effective typing is about equal to 5 hours, then that is 5 hours x 3,000 words/hour x = 15,000 words per day. This "Clubhouse-like" application, in my estimation should not take more than two days to program at this speed for a really good C++ programmer.

I see a potential for "No one types that fast!" I type 50 or more words per minute. I did. I am retired and I still do. Therefore, that statement of "No one" is wrong.

I see a potential for "No one is that good of a programmer!" Bjorne Stroustrup, and most or all of the members of the ISO that are responsible to analyze and pretend to stamp validation on his code probably are that good of programmers. Therefore, that statement of "No one" is probably wrong.

In my original response I said, "a high quality coder". I meant that. Not someone that is still learning and having to look up definitions and re-study Stroustrup's books, like me (I am still learning). But "a high quality coder". I expected it to take no more than 2 days to do this. But, I used to be a senior project manager at a structural steel company and I found that when the owners said that they had high quality structural steel fabricators that I often had to go to their shop and instruct some of the fabricators in how to be accurate. Thus, " in about 1 to 3 weeks" was my response instead of the actual maximum of 2 days that it should take. Remember, 15,000 words per day. Some people really do know how to program in C++ without having to look commands up in a reference book. Some people might not need to spend hours googling examples of how to apply those commands, like I need to. OK?

Cost for 1 to 3 weeks: A high quality coder might get $60,000 to $250,000 per year. WHAT? $60,000? "No one would..." Yes, they would. In deep south, back-woods Mississippi, or back-woods Canada, a programmer might not have the economic locals available to make more then $30,000 a year. Economics does not always stop a person from learning. Pay does not make better. I have met systems developers in back-country shops that I considered to be worth at least $200,000 more per year if they were working in software-valley California. One person that I met made about $7.50 per hour while the company that they worked for charged $150.00 per hour. I said, "If you are paying someone else to write it for you", meaning (in this case) that YOU hire them. Back to costs; 2 days, 16 hours total maybe. Chill out! It can be done by "a high quality coder". If that coder is accustomed to making $60,000 per year, and if you round down the workable hours to 2,000 per year, then that is $30.00 per hour. 2 days is 16 hours. 16 hours is $480.00 for the complete programming of a "Clubhouse-like app". Not a full, online, multi-server driven, internet-backbone sitting, very high bandwidth capacity connecting, system. Just a "Clubhouse-like app".

That is what the Original Poster requested. I stayed with that request. I answered that request.

You might wonder where I got the "Cost to code: Maybe around $9,000 to 30,000" when I expect it might only cost less than $500.00 for "a high quality coder" to program this. Experience at estimating in industry, and seeing how sloppy hiring practices can cost a company so severely much more than they should have had to pay. I have meet lots of people that are quite knowledgeable at some things. They might work at a Shell or Exxon gas station and in their free time have spent years learning something as an interest and have not really been considering how much money that they could be making at it. They might work at a cut-rate shoe store and have learned, as an interest, how to de-compile any and all Visual Studio and .net and ActiveX programs simply because they want to force the current virtual-3D, First-Person-Shooter, game that they are playing to be different. Such people exist. But, it is common in industry to have some know-less, self-deluded, hiring person that knows almost nothing about the skill set that they are filtering for, to make the initial decision as to who gets the job. So, if I think that it should take 2 days, then I say at least one week. And, if I think that one week is the minimum, then I say maybe 3 weeks. Experience. High quality welders should be interviewing and testing applicants for welding jobs. High quality tandom-rig drivers should be interviewing and testing applicants for Over The Road drivers. High quality programmers should be interviewing and testing applicants for programming jobs. But, this does not seem to be how a lot of upper level management thinks. They seem to think that some know-less with a master's degree in Human Resources can pick out the best programmer. I think that Stroustrup could probably code this in about 1 hour or less. I expect that of the ISO members that study what he sends to them. So, I adjusted that 1 hour to 1 day, then I doubled that 1 day to 2 days, then I was comfortable with it under the condition of "a high quality coder" that YOU hire, then when I considered that you and most readers might have very little idea how to hire "a high quality coder" I buffered this to 1 to 3 weeks.

If it took more than 2 hours for a programmer to set up a very basic, barely working, "Clubhouse-like" program in C++, then I would hand them their check for the day and allow them to go find somewhere else to work. A graphical user interface with protocol for file transfers would be enough to keep their job. It should take them less than 30 minutes, but I would give them 2 hours. I am not you, so I told you 1 to 3 weeks and $9,000 to 30,000. Anything past 1 week and you should lay-off or fire your HR personnel that were responsible for hiring that programmer, that includes yourself. If you were responsible, then you have a problem.

Think.

Here is one of various ways to find such a programmer: Every store that you go to, and every meeting that you are in, and every time you go to a cowboy rodeo, and every time that you sit next to someone or meet someone or see someone on a bus (where you are riding), ask them if they can program a network system in C++. Expect lots of people to tell you that you are stupid. Expect lots of ridicule. BUT, expect to find (among the pretenders that say that they can) a high quality coder that can program a network system in C++. It works. It should take less than 1 week to code in a "Clubhouse-like" app. It might take you days or years to find such a person, but you will probably end up with a list of personnel that can program some other amazing things fast also.

You have to do the work to find them. YOU have to.

[ivansav]

Approximately, the costs of clubhouse-like app development will start from $10,000-$30,000. The actual cost will depend on various factors like detailed specifications of the project, the number of specialists, and the desired functionality. Also, the complexity will influence the budget: an MVP will require fewer development hours than a full app with all the possible functions.

[solviostech]

According to the web source, developing an application like Clubhouse can cost around $30,000 - $40,000 for a solitary stage (Android or iOS) with every one of the fundamental highlights as it were. Nonetheless, by including progressed highlights and fostering the application for a considerable length of time, the cost can ascend to $90,000.

[Djohnavid021]

The development process may take anywhere from $50,000 to $100,000 depending on the complexity and time frame. The costs are based on a fixed product state, which means they don't include any hypothesis testing, which can increase the total price. The process can also take longer than expected, and the project manager will likely have to supervise the work. Listed below are some considerations when hiring a developer for this type of project.

[PollyR]

And how much will the annual support of such an application cost?

[Mia White]

Hey,
the basic Clubhouse-like app may cost you $29,000, and this price may get bigger depending on how advanced you want your app to be.