5 ms·

Thought experiment: how much money and how long would it take you to build a product with complete feature parity to Slack?

c-smile 9y ago
I personally can do UI of Slack client in 5-8 months (so 5-8 man/months) + one/two backend developers needed. And it will be almost native client with client distribution less than 5mb. As a proof, it took me 2 months to make this application: https://notes.sciter.com/ https://notes.sciter.com/ (and site for it in that time too).
- askafriend 9y ago
  That's the easy part. The hard part is scaling, marketing, and building a business/platform around the product so that you take it worldwide. Several people have built Slack clones and there's even multiple open source projects out there. Building a company and business at a scale like Slack is the incredibly difficult part and nobody has managed to replicate that.
  - c-smile 9y ago
    That's true. My personal educated estimation is that only 5-10% of income is spent on R&D in established companies.
- ctraganos 9y ago
  I'd pay for an alternative "lite" Slack client for basic channels, DMs, and attachments. With 10+ accounts, my 2017 mac struggles under the weight of the Electron mac client
itomato 9y ago
A competitive app would have: - easy signup for public access - integrations for killer apps in other verticals - plugin architecture - public and private apis - mobile app(s) with 80% of core desktop functionality - end-to-end encryption Maybe it's sufficient to extend IRC. Add support for custom emojis & Giphy integration. Voice, video conferencing and screenshare would be essential.
- c-smile 9y ago
  I've participated in design of application similar to that (it is already on the market): voice, video, camera desktop sharing and textual chat. First version was made in 9..11 months: core UI team - 3 developers for desktop. 2 developers for mobile clients, one for each platform. Server side 3-5 as it supports video transmission for 1000 and more participants. Platforms: Windows, Mac OS, Linux, Android and iOS. Small, manageable and so effective professional team.
uberemployee 9y ago
Uber was on Hipchat, briefly experimented with Slack, went back to Hipchat, and then moved to our own Inhouse chat based on open source project, starting around the beginning the year. It was bad the first few months, but except for a recently publicized utterly embarrassing security vulnerability, in the last 2 months it's been working great. The mobile app is now much, much better than the HipChat ever was.
- wikibob 9y ago
  What project is the internal tool based on? Mattermost?
  - hkarthik 9y ago
    Yes, the internal tool is called uChat and it's based on Mattermost but Uber heavily contributed to Mattermost to make it scale. More details here: https://eng.uber.com/uchat/ https://eng.uber.com/uchat/
NelsonMinar 9y ago
Ask Discord.

c-smile 9y ago

I personally can do UI of Slack client in 5-8 months (so 5-8 man/months) + one/two backend developers needed. And it will be almost native client with client distribution less than 5mb. As a proof, it took me 2 months to make this application: https://notes.sciter.com/ https://notes.sciter.com/ (and site for it in that time too).

askafriend 9y ago
That's the easy part. The hard part is scaling, marketing, and building a business/platform around the product so that you take it worldwide. Several people have built Slack clones and there's even multiple open source projects out there. Building a company and business at a scale like Slack is the incredibly difficult part and nobody has managed to replicate that.
- c-smile 9y ago
  That's true. My personal educated estimation is that only 5-10% of income is spent on R&D in established companies.
ctraganos 9y ago
I'd pay for an alternative "lite" Slack client for basic channels, DMs, and attachments. With 10+ accounts, my 2017 mac struggles under the weight of the Electron mac client

itomato 9y ago

A competitive app would have: - easy signup for public access - integrations for killer apps in other verticals - plugin architecture - public and private apis - mobile app(s) with 80% of core desktop functionality - end-to-end encryption Maybe it's sufficient to extend IRC. Add support for custom emojis & Giphy integration. Voice, video conferencing and screenshare would be essential.

c-smile 9y ago
I've participated in design of application similar to that (it is already on the market): voice, video, camera desktop sharing and textual chat. First version was made in 9..11 months: core UI team - 3 developers for desktop. 2 developers for mobile clients, one for each platform. Server side 3-5 as it supports video transmission for 1000 and more participants. Platforms: Windows, Mac OS, Linux, Android and iOS. Small, manageable and so effective professional team.

uberemployee 9y ago

Uber was on Hipchat, briefly experimented with Slack, went back to Hipchat, and then moved to our own Inhouse chat based on open source project, starting around the beginning the year. It was bad the first few months, but except for a recently publicized utterly embarrassing security vulnerability, in the last 2 months it's been working great. The mobile app is now much, much better than the HipChat ever was.

wikibob 9y ago
What project is the internal tool based on? Mattermost?
- hkarthik 9y ago
  Yes, the internal tool is called uChat and it's based on Mattermost but Uber heavily contributed to Mattermost to make it scale. More details here: https://eng.uber.com/uchat/ https://eng.uber.com/uchat/

NelsonMinar 9y ago

Ask Discord.